Merge remote-tracking branch 'origin/master' into fix/ppi-kaks-enable

# Conflicts:
#	tests/unit/test_router_analysis_handoff_boundary.py

.gitignore

@@ -18,7 +18,7 @@ outputs/
 input/
 workspace/
 paper/
-memory/
+#memory/
 memory_storage/
 backup/
 backups/
@@ -29,3 +29,4 @@ nohup.out
 Rplots.pdf
 *.pyc
 *.log
+.gitignore

apps/cli/run_gene_family_report.py

@@ -0,0 +1,70 @@
+#!/usr/bin/env python3
+"""CLI for the plan-scoped formal gene-family result report."""
+
+from __future__ import annotations
+
+import argparse
+import json
+from pathlib import Path
+import sys
+from typing import Sequence
+
+PROJECT_ROOT = Path(__file__).resolve().parents[2]
+if str(PROJECT_ROOT) not in sys.path:
+    sys.path.insert(0, str(PROJECT_ROOT))
+
+from gfa_core.gene_family_report import GeneFamilyReportResult, generate_gene_family_report
+
+
+def _parser() -> argparse.ArgumentParser:
+    parser = argparse.ArgumentParser(description="Generate or open a gene-family result report")
+    parser.add_argument("--workspace", type=Path, required=True)
+    parser.add_argument("--prefix", required=True)
+    parser.add_argument("--job-id", default="")
+    parser.add_argument("--expected-stages", default="")
+    parser.add_argument("--stage-statuses-json", default="{}")
+    parser.add_argument("--scope-id", default="")
+    parser.add_argument("--analysis-route", default="")
+    parser.add_argument("--member-set-id", default="")
+    parser.add_argument("--rnaseq-scope-id", default="")
+    parser.add_argument("--action", choices=("generate", "open"), default="generate")
+    parser.add_argument("--no-model", action="store_true")
+    return parser
+
+
+def main(argv: Sequence[str] | None = None) -> GeneFamilyReportResult:
+    args = _parser().parse_args(argv)
+    statuses = json.loads(args.stage_statuses_json)
+    if not isinstance(statuses, dict):
+        raise ValueError("--stage-statuses-json must contain a JSON object")
+    result = generate_gene_family_report(
+        args.workspace,
+        args.prefix,
+        expected_stages=[value.strip() for value in args.expected_stages.split(",") if value.strip()],
+        stage_statuses={str(key): str(value) for key, value in statuses.items()},
+        scope_id=args.scope_id,
+        analysis_route=args.analysis_route,
+        member_set_id=args.member_set_id,
+        rnaseq_scope_id=args.rnaseq_scope_id,
+        use_model=not args.no_model,
+        action=args.action,
+        job_id=args.job_id,
+    )
+    print(
+        json.dumps(
+            {
+                "status": "success",
+                "report_path": str(result.report_path),
+                "manifest_path": str(result.manifest_path),
+                "report_status": result.report_status,
+                "generation_mode": result.generation_mode,
+                "reused": result.reused,
+            },
+            ensure_ascii=False,
+        )
+    )
+    return result
+
+
+if __name__ == "__main__":
+    main()

apps/web/pipeline_backend.py

@@ -6112,8 +6112,9 @@ def run_stage(
 
     target_species = kwargs.get("target_species")
     interspecies_ref_species = kwargs.get("interspecies_ref_species")
-    scope_file = str(kwargs.get("scope_file") or kwargs.get("scopeFile") or "").strip()
-    analysis_route = str(kwargs.get("analysis_route") or kwargs.get("analysisRoute") or "").strip()
+    scope_file = str(kwargs.get("scope_file") or kwargs.get("scopeFile") or "").strip()
+    scope_id = str(kwargs.get("scope_id") or kwargs.get("scopeId") or "").strip()
+    analysis_route = str(kwargs.get("analysis_route") or kwargs.get("analysisRoute") or "").strip()
     analysis_mode = str(kwargs.get("analysis_mode") or kwargs.get("analysisMode") or "").strip()
     member_set_id = str(kwargs.get("member_set_id") or kwargs.get("memberSetId") or "").strip()
     combo_output_label = str(combo_output_label or kwargs.get("comboOutputLabel") or kwargs.get("combo_output_label") or "").strip()
@@ -8432,7 +8433,62 @@ def run_stage(
         else:
             return f"❌ PEP failed (exit code: {code}). Log: {log_file}", manifest
 
-    if stage == "summary":
+    if stage == "gene_family_report":
+        expected_stages = kwargs.get("expected_stages") or kwargs.get("expectedStages") or []
+        if isinstance(expected_stages, str):
+            expected_stages = [value.strip() for value in expected_stages.split(",") if value.strip()]
+        else:
+            expected_stages = [str(value).strip() for value in expected_stages if str(value).strip()]
+        report_action = str(kwargs.get("report_action") or kwargs.get("reportAction") or "generate").strip()
+        if report_action == "generate" and not expected_stages:
+            return "❌ 基因家族结果报告生成需要非空 expected_stages", manifest
+        stage_statuses = kwargs.get("stage_statuses") or kwargs.get("stageStatuses") or {}
+        if isinstance(stage_statuses, str):
+            try:
+                stage_statuses = json.loads(stage_statuses)
+            except json.JSONDecodeError:
+                return "❌ 基因家族结果报告的 stage_statuses 不是合法 JSON", manifest
+        if not isinstance(stage_statuses, dict):
+            return "❌ 基因家族结果报告的 stage_statuses 必须是对象", manifest
+        report_use_model_raw = kwargs.get("report_use_model", kwargs.get("reportUseModel", True))
+        rnaseq_scope_id = str(kwargs.get("rnaseq_scope_id") or kwargs.get("rnaseqScopeId") or "").strip()
+        cmd = [
+            sys.executable,
+            "apps/cli/run_gene_family_report.py",
+            "--workspace",
+            str(ws),
+            "--prefix",
+            prefix,
+            "--job-id",
+            job.job_id,
+            "--expected-stages",
+            ",".join(expected_stages),
+            "--stage-statuses-json",
+            json.dumps(stage_statuses, ensure_ascii=False, separators=(",", ":")),
+            "--action",
+            report_action,
+        ]
+        for flag, value in (
+            ("--scope-id", scope_id),
+            ("--analysis-route", analysis_route),
+            ("--member-set-id", member_set_id),
+            ("--rnaseq-scope-id", rnaseq_scope_id),
+        ):
+            if value:
+                cmd += [flag, value]
+        if not _truthy(report_use_model_raw):
+            cmd.append("--no-model")
+        code = run_cmd_logged(cmd, log_file, cwd=Path("."))
+        report_manifests = sorted(
+            final_dir.glob(f"{prefix}*.gene_family_report.manifest.json"),
+            key=lambda path: path.stat().st_mtime,
+        )
+        report_manifest = report_manifests[-1] if report_manifests else manifest
+        if code != 0:
+            return f"❌ 基因家族结果报告生成失败，日志：{log_file}", report_manifest
+        return f"✅ 基因家族结果报告已生成。日志：{log_file}", report_manifest
+
+    if stage == "summary":
         summary_manifest_seed = (
             ws / "final" / f"{prefix}.identify.manifest.json"
             if (ws / "final" / f"{prefix}.identify.manifest.json").exists()

apps/web/vue3/agent_status_component_ml.html

@@ -0,0 +1,416 @@
+<!-- Agent状态显示组件 - 可嵌入index.html -->
+<template id="agent-status-component">
+  <div class="agent-status-panel">
+    <div class="agent-header">
+      <h3>{{ t('agent-assistant') }}</h3>
+      <span class="agent-mode">{{ currentMode }}</span>
+    </div>
+    
+    <div class="agent-messages" ref="messageContainer">
+      <div v-for="(msg, index) in messages" :key="index" 
+           :class="['message', msg.role]">
+        <div class="message-role">{{ getRoleLabel(msg.role) }}</div>
+        <div class="message-content" v-html="formatMessage(msg.content)"></div>
+        <div class="message-time">{{ formatTime(msg.timestamp) }}</div>
+      </div>
+    </div>
+    
+    <div class="agent-input">
+      <textarea v-model="userInput" 
+                :placeholder="t('enter-question')"
+                @keydown.enter.prevent="sendMessage"
+                rows="3"></textarea>
+      <button @click="sendMessage" :disabled="isProcessing">
+        {{ isProcessing ? t('processing') : t('send') }}
+      </button>
+    </div>
+    
+    <div class="agent-tasks" v-if="tasks.length > 0">
+      <h4>{{ t('task-list') }}</h4>
+      <div v-for="task in tasks" :key="task.task_id" 
+           :class="['task-item', task.status]"
+           @click="selectTask(task)">
+        <span class="task-status">{{ getStatusIcon(task.status) }}</span>
+        <span class="task-message">{{ task.message }}</span>
+        <span class="task-progress">{{ task.progress }}%</span>
+      </div>
+    </div>
+  </div>
+</template>
+
+<script src="/static/i18n.js"></script>
+// Agent状态组件
+const AgentStatusComponent = {
+  template: '#agent-status-component',
+  
+  data() {
+    return {
+      messages: [],
+      userInput: '',
+      isProcessing: false,
+      currentMode: '',
+      sessionId: '',
+      tasks: [],
+      currentTask: null,
+      pollingInterval: null
+    };
+  },
+  
+  mounted() {
+    this.sessionId = 'session_' + Date.now();
+    this.currentMode = t('standby');
+    this.loadTasks();
+    this.startPolling();
+  },
+  
+  beforeUnmount() {
+    this.stopPolling();
+  },
+  
+  methods: {
+    async sendMessage() {
+      if (!this.userInput.trim() || this.isProcessing) return;
+      
+      const message = this.userInput.trim();
+      this.userInput = '';
+      this.isProcessing = true;
+      
+      // 添加用户消息
+      this.addMessage('user', message);
+      
+      try {
+        // 调用Agent执行API
+        const response = await fetch('/api/agent/execute', {
+          method: 'POST',
+          headers: { 'Content-Type': 'application/json' },
+          body: JSON.stringify({
+            message: message,
+            session_id: this.sessionId,
+            model: 'kimi'
+          })
+        });
+        
+        const data = await response.json();
+        
+        if (data.success) {
+          this.currentTask = data.task_id;
+          this.addMessage('system', t('msg-task-submitted') + data.task_id);
+          this.loadTasks();
+        } else {
+          this.addMessage('error', data.message || t('msg-submit-failed'));
+        }
+      } catch (error) {
+        this.addMessage('error', t('msg-request-failed') + error.message);
+      }
+      
+      this.isProcessing = false;
+    },
+    
+    async loadTasks() {
+      try {
+        const response = await fetch('/api/agent/tasks');
+        const data = await response.json();
+        
+        if (data.success) {
+          this.tasks = data.tasks;
+        }
+      } catch (error) {
+        console.error(t('msg-load-task-failed'), error);
+      }
+    },
+    
+    async checkTaskStatus(taskId) {
+      try {
+        const response = await fetch(`/api/agent/status/${taskId}`);
+        const data = await response.json();
+        
+        if (data.success && data.task) {
+          const task = data.task;
+          
+          // 更新任务状态
+          const taskIndex = this.tasks.findIndex(t => t.task_id === taskId);
+          if (taskIndex >= 0) {
+            this.tasks[taskIndex] = { ...this.tasks[taskIndex], ...task };
+          }
+          
+          // 如果任务完成，显示结果
+          if (task.status === 'completed' && task.result) {
+            this.addMessage('assistant', task.result.response || t('msg-task-completed'));
+            
+            // 显示执行结果摘要
+            if (task.result.execution_result) {
+              const execResult = task.result.execution_result;
+              if (execResult.status === 'success') {
+                const manifest = execResult.manifest || {};
+                const metrics = manifest.metrics || {};
+                this.addMessage('system',
+                  t('msg-analysis-complete') + (metrics.final_members || 0) + t('msg-gene-members'));
+              }
+            }
+          } else if (task.status === 'failed') {
+            this.addMessage('error', task.result?.response || t('msg-task-failed'));
+          }
+        }
+      } catch (error) {
+        console.error(t('msg-check-task-status-failed'), error);
+      }
+    },
+    
+    selectTask(task) {
+      this.currentTask = task.task_id;
+      this.checkTaskStatus(task.task_id);
+    },
+    
+    startPolling() {
+      this.pollingInterval = setInterval(() => {
+        if (this.currentTask) {
+          this.checkTaskStatus(this.currentTask);
+        }
+        this.loadTasks();
+      }, 3000); // 每3秒轮询一次
+    },
+    
+    stopPolling() {
+      if (this.pollingInterval) {
+        clearInterval(this.pollingInterval);
+      }
+    },
+    
+    addMessage(role, content) {
+      this.messages.push({
+        role: role,
+        content: content,
+        timestamp: new Date().toISOString()
+      });
+      
+      // 滚动到底部
+      setTimeout(() => {
+        const container = this.$refs.messageContainer;
+        if (container) {
+          container.scrollTop = container.scrollHeight;
+        }
+      }, 100);
+    },
+    
+    getRoleLabel(role) {
+      const keys = {
+        'user': 'user-role',
+        'assistant': 'agent-role',
+        'system': 'system-role',
+        'error': 'error-role',
+        'execution': 'execution-role',
+        'planning': 'planning-role',
+        'diagnosis': 'diagnosis-role'
+      };
+      return keys[role] ? t(keys[role]) : role;
+    },
+    
+    getStatusIcon(status) {
+      const icons = {
+        'pending': '⏳',
+        'running': '🔄',
+        'completed': '✅',
+        'failed': '❌',
+        'cancelled': '🚫'
+      };
+      return icons[status] || '❓';
+    },
+    
+    formatMessage(content) {
+      // 简单的Markdown格式化
+      return content
+        .replace(/\*\*(.+?)\*\*/g, '<strong>$1</strong>')
+        .replace(/\n/g, '<br>');
+    },
+    
+    formatTime(timestamp) {
+      if (!timestamp) return '';
+      const date = new Date(timestamp);
+      return date.toLocaleTimeString('zh-CN', {
+        hour: '2-digit',
+        minute: '2-digit',
+        second: '2-digit'
+      });
+    },
+
+    t(key) {
+      return window.t ? window.t(key) : key;
+    }
+  }
+};
+</script>
+
+<style>
+.agent-status-panel {
+  border: 1px solid #ddd;
+  border-radius: 8px;
+  padding: 16px;
+  background: #f9f9f9;
+  height: 100%;
+  display: flex;
+  flex-direction: column;
+}
+
+.agent-header {
+  display: flex;
+  justify-content: space-between;
+  align-items: center;
+  margin-bottom: 12px;
+  padding-bottom: 8px;
+  border-bottom: 1px solid #eee;
+}
+
+.agent-header h3 {
+  margin: 0;
+  font-size: 16px;
+}
+
+.agent-mode {
+  font-size: 12px;
+  color: #666;
+  background: #e0e0e0;
+  padding: 2px 8px;
+  border-radius: 4px;
+}
+
+.agent-messages {
+  flex: 1;
+  overflow-y: auto;
+  max-height: 400px;
+  margin-bottom: 12px;
+}
+
+.message {
+  margin-bottom: 12px;
+  padding: 8px 12px;
+  border-radius: 8px;
+  background: white;
+  border-left: 3px solid #ddd;
+}
+
+.message.user {
+  border-left-color: #4CAF50;
+  background: #E8F5E9;
+}
+
+.message.assistant {
+  border-left-color: #2196F3;
+  background: #E3F2FD;
+}
+
+.message.error {
+  border-left-color: #f44336;
+  background: #FFEBEE;
+}
+
+.message.system {
+  border-left-color: #9E9E9E;
+  background: #F5F5F5;
+  font-size: 12px;
+}
+
+.message-role {
+  font-size: 12px;
+  color: #666;
+  margin-bottom: 4px;
+}
+
+.message-content {
+  font-size: 14px;
+  line-height: 1.5;
+}
+
+.message-time {
+  font-size: 11px;
+  color: #999;
+  text-align: right;
+  margin-top: 4px;
+}
+
+.agent-input {
+  display: flex;
+  gap: 8px;
+  margin-bottom: 12px;
+}
+
+.agent-input textarea {
+  flex: 1;
+  padding: 8px;
+  border: 1px solid #ddd;
+  border-radius: 4px;
+  resize: none;
+  font-family: inherit;
+}
+
+.agent-input button {
+  padding: 8px 16px;
+  background: #2196F3;
+  color: white;
+  border: none;
+  border-radius: 4px;
+  cursor: pointer;
+}
+
+.agent-input button:disabled {
+  background: #ccc;
+  cursor: not-allowed;
+}
+
+.agent-tasks {
+  border-top: 1px solid #eee;
+  padding-top: 12px;
+}
+
+.agent-tasks h4 {
+  margin: 0 0 8px 0;
+  font-size: 14px;
+}
+
+.task-item {
+  display: flex;
+  align-items: center;
+  gap: 8px;
+  padding: 6px 8px;
+  border-radius: 4px;
+  cursor: pointer;
+  margin-bottom: 4px;
+  background: white;
+}
+
+.task-item:hover {
+  background: #f0f0f0;
+}
+
+.task-item.pending {
+  border-left: 3px solid #FFC107;
+}
+
+.task-item.running {
+  border-left: 3px solid #2196F3;
+}
+
+.task-item.completed {
+  border-left: 3px solid #4CAF50;
+}
+
+.task-item.failed {
+  border-left: 3px solid #f44336;
+}
+
+.task-status {
+  font-size: 14px;
+}
+
+.task-message {
+  flex: 1;
+  font-size: 12px;
+  overflow: hidden;
+  text-overflow: ellipsis;
+  white-space: nowrap;
+}
+
+.task-progress {
+  font-size: 11px;
+  color: #666;
+}
+</style>

apps/web/vue3/index-with-sidebar_ml.html

@@ -0,0 +1,391 @@
+<!DOCTYPE html>
+<html lang="zh-CN">
+<head>
+    <meta charset="UTF-8">
+    <script src="/static/i18n.js"></script>
+    <meta name="viewport" content="width=device-width, initial-scale=1.0">
+    <title>BotanicalBaseAgent</title>
+    <link rel="stylesheet" href="https://cdn.jsdelivr.net/npm/element-plus@2.3.14/dist/index.css">
+    <script src="https://cdn.jsdelivr.net/npm/vue@3.3.4/dist/vue.global.prod.js"></script>
+    <script src="https://cdn.jsdelivr.net/npm/element-plus@2.3.14/dist/index.full.js"></script>
+    <script src="https://cdn.jsdelivr.net/npm/@element-plus/icons-vue@2.1.0/dist/index.iife.js"></script>
+    <link rel="stylesheet" href="conversation-sidebar.css">
+    <style>
+        * { box-sizing: border-box; }
+        body { 
+            margin: 0; 
+            font-family: 'Helvetica Neue', Arial, sans-serif; 
+            background: #f5f7fa; 
+        }
+        .header { 
+            background: linear-gradient(135deg, #667eea 0%, #764ba2 100%); 
+            color: white; 
+            padding: 15px 20px; 
+        }
+        .header h1 { margin: 0; font-size: 22px; }
+        .main-container { 
+            display: flex; 
+            height: calc(100vh - 60px); 
+            overflow: hidden;
+        }
+        
+        /* 左侧控制面板 */
+        .left-panel {
+            width: 280px;
+            background: white;
+            border-right: 1px solid #e4e7ed;
+            display: flex;
+            flex-direction: column;
+            padding: 15px;
+            gap: 15px;
+            overflow-y: auto;
+        }
+        
+        /* 中间内容区 */
+        .center-panel {
+            flex: 1;
+            display: flex;
+            flex-direction: column;
+            padding: 15px;
+            gap: 15px;
+            overflow: hidden;
+        }
+        
+        /* 右侧结果区 */
+        .right-panel {
+            width: 350px;
+            background: white;
+            border-left: 1px solid #e4e7ed;
+            padding: 15px;
+            overflow-y: auto;
+        }
+        
+        .panel-section {
+            background: white;
+            border-radius: 8px;
+            padding: 15px;
+            box-shadow: 0 2px 8px rgba(0,0,0,0.08);
+        }
+        
+        .panel-title {
+            font-size: 14px;
+            font-weight: bold;
+            color: #409EFF;
+            margin-bottom: 12px;
+            padding-bottom: 8px;
+            border-bottom: 2px solid #409EFF;
+        }
+        
+        .chat-area {
+            flex: 1;
+            background: white;
+            border-radius: 8px;
+            padding: 15px;
+            overflow-y: auto;
+            box-shadow: 0 2px 8px rgba(0,0,0,0.08);
+        }
+        
+        .chat-message {
+            margin-bottom: 15px;
+            padding: 12px 16px;
+            border-radius: 8px;
+            max-width: 80%;
+        }
+        
+        .chat-message.user {
+            background: #ecf5ff;
+            margin-left: auto;
+            text-align: right;
+        }
+        
+        .chat-message.assistant {
+            background: #f0f9eb;
+        }
+        
+        .input-area {
+            display: flex;
+            gap: 10px;
+        }
+        
+        .stage-tag {
+            display: inline-block;
+            padding: 4px 12px;
+            border-radius: 4px;
+            font-size: 12px;
+            margin: 4px;
+        }
+        
+        .stage-tag.success { background: #67C23A; color: white; }
+        .stage-tag.running { background: #E6A23C; color: white; }
+        .stage-tag.pending { background: #909399; color: white; }
+    </style>
+</head>
+<body>
+    <div id="app">
+        <!-- Header -->
+        <div class="header">
+            <h1>{{ t('gf-analysis-system') }}</h1>
+        </div>
+
+        <div class="main-container">
+            <!-- 左侧：会话侧边栏 + 控制面板 -->
+            <div class="left-panel">
+                <!-- 会话侧边栏 -->
+                <div class="conversation-sidebar" style="width: 100%; margin: -15px; margin-bottom: 0;">
+                    <div class="sidebar-header">
+                        <button class="new-chat-btn" @click="createNewConversation" style="flex: 1;">
+                            <el-icon><plus /></el-icon>
+                            <span>New Chat</span>
+                        </button>
+                    </div>
+                    
+                    <div class="search-box" style="padding: 12px;">
+                        <el-input 
+                            v-model="searchQuery" 
+                            :placeholder="t('search-conv')"
+                            prefix-icon="search" 
+                            clearable 
+                            size="small">
+                        </el-input>
+                    </div>
+                    
+                    <div class="conversation-list" style="max-height: 300px;">
+                        <div class="list-section">
+                            <div class="section-title">{{ t('conv-list') }}</div>
+                            <div v-for="conv in filteredConversations" 
+                                 :key="conv.id"
+                                 class="conversation-item" 
+                                 :class="{ active: currentConv?.id === conv.id }"
+                                 @click="selectConversation(conv)">
+                                <div class="item-icon">
+                                    <el-icon><chat-dot-round /></el-icon>
+                                </div>
+                                <div class="item-content">
+                                    <div class="item-title">{{ conv.title }}</div>
+                                    <div class="item-meta">
+                                        <span class="gene-family" v-if="conv.gene_family">{{ conv.gene_family }}</span>
+                                        <span class="message-count">{{ conv.message_count }}{{ t('msg-count') }}</span>
+                                    </div>
+                                </div>
+                            </div>
+                        </div>
+                    </div>
+                </div>
+                
+                <!-- 分析控制 -->
+                <div class="panel-section">
+                    <div class="panel-title">{{ t('analysis-stage') }}</div>
+                    <el-select v-model="selectedStage" :placeholder="t('select-stage')" style="width: 100%;">
+                        <el-option :label="t('identify-label')" value="identify"></el-option>
+                        <el-option :label="t('phylogeny-label')" value="phylogeny"></el-option>
+                        <el-option :label="t('motif-label')" value="motif"></el-option>
+                        <el-option :label="t('domain-label')" value="domain"></el-option>
+                        <el-option :label="t('expression-label')" value="expression"></el-option>
+                    </el-select>
+                    <el-button type="primary" style="width: 100%; margin-top: 10px;" @click="executeAnalysis">
+                        {{ t('execute-analysis') }}
+                    </el-button>
+                </div>
+
+                <!-- 参数设置 -->
+                <div class="panel-section">
+                    <div class="panel-title">{{ t('param-settings') }}</div>
+                    <el-form :model="params" label-width="80px" size="small">
+                        <el-form-item :label="t('prefix')">
+                            <el-input v-model="params.prefix" :placeholder="t('prefix-placeholder')"></el-input>
+                        </el-form-item>
+                        <el-form-item label="CPU">
+                            <el-input-number v-model="params.cpu" :min="1" :max="32" style="width: 100%;"></el-input-number>
+                        </el-form-item>
+                    </el-form>
+                </div>
+            </div>
+            
+            <!-- 中间：对话区 -->
+            <div class="center-panel">
+                <div class="chat-area">
+                    <div v-for="(msg, index) in messages" :key="index" 
+                         class="chat-message" :class="msg.role">
+                        <div style="font-size: 12px; color: #666; margin-bottom: 4px;">
+                            {{ msg.role === 'user' ? t('user') : t('ai-assistant') }} · {{ msg.time }}
+                        </div>
+                        <div>{{ msg.content }}</div>
+                    </div>
+                    <div v-if="messages.length === 0" style="text-align: center; color: #999; padding: 40px;">
+                        <el-icon :size="48" style="display: block; margin-bottom: 10px;"><chat-line-round /></el-icon>
+                        <p>{{ t('start-new-chat') }}</p>
+                        <p style="font-size: 12px;">{{ t('or-continue') }}</p>
+                    </div>
+                </div>
+                
+                <div class="input-area">
+                    <el-input 
+                        v-model="userInput" 
+                        type="textarea" 
+                        :rows="3" 
+                        :placeholder="t('describe-need')"
+                        @keyup.enter.ctrl="sendMessage">
+                    </el-input>
+                    <el-button type="primary" size="large" @click="sendMessage" :disabled="!userInput.trim()">
+                        {{ t('send') }}
+                    </el-button>
+                </div>
+            </div>
+            
+            <!-- 右侧：结果展示 -->
+            <div class="right-panel">
+                <div class="panel-title">📊 Analysis Results</div>
+                
+                <div v-for="stage in stages" :key="stage.name" style="margin-bottom: 15px;">
+                    <div style="display: flex; justify-content: space-between; align-items: center; margin-bottom: 8px;">
+                        <span style="font-weight: 500;">{{ stage.label }}</span>
+                        <span class="stage-tag" :class="stage.status">{{ stage.statusText }}</span>
+                    </div>
+                    <el-progress :percentage="stage.progress" :status="stage.status"></el-progress>
+                </div>
+                
+                <el-divider></el-divider>
+                
+                <div style="text-align: center; color: #999; padding: 20px;">
+                    <el-icon :size="48" style="display: block; margin-bottom: 10px;"><picture /></el-icon>
+                    <p>{{ t('results-here') }}</p>
+                </div>
+            </div>
+        </div>
+    </div>
+
+    <script>
+        const { createApp, ref, reactive, computed } = Vue;
+        
+        const app = createApp({
+            setup() {
+                const t = window.t;
+                document.title = 'BotanicalBaseAgent - ' + t('gf-analysis');
+                // 会话管理
+                const conversations = ref([
+                    { id: 1, title: 'Ft.SPL基因家族分析', gene_family: 'SPL', message_count: 12 },
+                    { id: 2, title: 'Os.TCP进化分析', gene_family: 'TCP', message_count: 5 },
+                    { id: 3, title: 'HMMER参数配置咨询', gene_family: null, message_count: 8 },
+                ]);
+                const currentConv = ref(null);
+                const searchQuery = ref('');
+                
+                const filteredConversations = computed(() => {
+                    if (!searchQuery.value) return conversations.value;
+                    const query = searchQuery.value.toLowerCase();
+                    return conversations.value.filter(conv => 
+                        conv.title.toLowerCase().includes(query) ||
+                        (conv.gene_family && conv.gene_family.toLowerCase().includes(query))
+                    );
+                });
+                
+                const createNewConversation = () => {
+                    const newConv = {
+                        id: Date.now(),
+                        title: t('new-chat') + ' ' + new Date().toLocaleTimeString(),
+                        gene_family: null,
+                        message_count: 0
+                    };
+                    conversations.value.unshift(newConv);
+                    selectConversation(newConv);
+                };
+                
+                const selectConversation = (conv) => {
+                    currentConv.value = conv;
+                    // 清空消息并添加欢迎语
+                    messages.value = [{
+                        role: 'assistant',
+                        content: t('initial-welcome').replace('{title}', conv.title),
+                        time: new Date().toLocaleTimeString()
+                    }];
+                };
+                
+                // 分析参数
+                const selectedStage = ref('');
+                const params = reactive({
+                    prefix: 'Ft.SPL',
+                    cpu: 8
+                });
+                
+                // 消息
+                const messages = ref([]);
+                const userInput = ref('');
+                
+                const sendMessage = () => {
+                    if (!userInput.value.trim()) return;
+                    
+                    // 添加用户消息
+                    messages.value.push({
+                        role: 'user',
+                        content: userInput.value,
+                        time: new Date().toLocaleTimeString()
+                    });
+                    
+                    // 更新当前会话消息数
+                    if (currentConv.value) {
+                        currentConv.value.message_count++;
+                    }
+                    
+                    // 模拟AI回复
+                    setTimeout(() => {
+                        messages.value.push({
+                            role: 'assistant',
+                            content: `Received: "${userInput.value}". I will help analyze the gene family. Confirm parameters and click "${t('execute-analysis')}".`,
+                            time: new Date().toLocaleTimeString()
+                        });
+                        if (currentConv.value) {
+                            currentConv.value.message_count++;
+                        }
+                    }, 500);
+                    
+                    userInput.value = '';
+                };
+                
+                // 分析阶段
+                const stages = ref([
+                    { name: 'identify', label: 'Identify', status: 'success', statusText: t('completed'), progress: 100 },
+                    { name: 'phylogeny', label: 'Phylogeny', status: 'running', statusText: t('running'), progress: 60 },
+                    { name: 'motif', label: 'Motif', status: 'pending', statusText: t('pending'), progress: 0 },
+                    { name: 'domain', label: 'Domain', status: 'pending', statusText: t('pending'), progress: 0 },
+                ]);
+                
+                const executeAnalysis = () => {
+                    if (!selectedStage.value) {
+                        ElementPlus.ElMessage.warning(t('msg-select-stage-first'));
+                        return;
+                    }
+                    ElementPlus.ElMessage.success(t('msg-executing') + selectedStage.value + t('msg-analysis-text'));
+                };
+                
+                return {
+                    t,
+                    conversations,
+                    currentConv,
+                    searchQuery,
+                    filteredConversations,
+                    createNewConversation,
+                    selectConversation,
+                    selectedStage,
+                    params,
+                    messages,
+                    userInput,
+                    sendMessage,
+                    stages,
+                    executeAnalysis
+                };
+            }
+        });
+        
+        app.use(ElementPlus);
+        // 注册 i18n 翻译函数到 Vue 模板作用域
+        app.config.globalProperties.t = window.t || ((key) => key);
+
+        // 注册图标
+        for (const [key, component] of Object.entries(ElementPlusIconsVue)) {
+            app.component(key, component);
+        }
+        
+        app.mount('#app');
+    </script>
+</body>
+</html>

apps/web/vue3/index.html

@@ -1019,7 +1019,7 @@
         }
 
         .execution-log-item.assistant .log-content.structured-content {
-            /* 结构化内容使用rag样式 */
+            font-size: 13px;
         }
 
         .execution-log-item.assistant .log-content.short-content {
@@ -1387,7 +1387,8 @@
 
                     <el-divider style="margin: 12px 0;"></el-divider>
 
-                    <div v-if="false" style="padding:10px; background:#fff; border:1px solid #dcdfe6; border-radius:0px;">
+                    <div @dragover.prevent @drop.prevent="handleLightweightExtensionFolderDrop($event)"
+                        style="padding:10px; background:#fff; border:1px dashed #409EFF; border-radius:0px;">
                         <div style="display:flex; align-items:center; justify-content:space-between; gap:8px; margin-bottom:8px;">
                             <el-tooltip content="上传已跑通过的代码和测试数据；执行通过自然语言触发。" placement="top">
                                 <h4 style="margin:0; color:#303133; cursor:help;">拓展分析阶段</h4>
@@ -1413,10 +1414,13 @@
                             <span>名称</span><el-input v-model="extensionBuild.displayName" size="small" placeholder="例如 HMMER hmmsearch"></el-input>
                             <span>原始命令</span><el-input v-model="extensionBuild.command" size="small" placeholder="例如 bash run.sh input.fa"></el-input>
                             <span>环境</span>
-                            <el-select v-model="extensionBuild.runtimeEnv" size="small">
-                                <el-option label="test4py311" value="test4py311"></el-option>
-                                <el-option label="rnaseq" value="rnaseq"></el-option>
-                                <el-option label="rnaR" value="rnaR"></el-option>
+                            <el-select v-model="extensionBuild.runtimeEnv" size="small" filterable>
+                                <el-option
+                                    v-for="env in extensionEnvOptions"
+                                    :key="env"
+                                    :label="env"
+                                    :value="env">
+                                </el-option>
                             </el-select>
                             <span>原始代码</span><input type="file" multiple @change="onExtensionCodeFilesChange" style="font-size:12px; min-width:0;">
                             <span v-if="extensionBuild.codeFiles.length"></span>
@@ -1801,6 +1805,20 @@
                             📂 Select
                         </el-button>
                     </div>
+                    <!-- lightWeightExtensionTest:start -->
+                    <div id="lightweight-extension-test-drop-zone"
+                        style="display:none;margin-top:8px;border:1px dashed #409EFF;padding:10px;font-size:12px;color:#606266;background:#f8fbff;"
+                         @dragover.prevent
+                         @drop.prevent="handleLightweightExtensionFolderDrop($event)">
+                        <div style="font-weight:600;color:#303133;">拖入固定结构拓展包</div>
+                        <div style="margin-top:4px;line-height:1.5;">要求包含 raw_code/run.sh 与 raw_test_data/，系统会自动创建 Job、解析环境并开始 lightweight upload。</div>
+                        <input id="lightweight-extension-test-folder-input" type="file" webkitdirectory directory multiple style="display:none" @change="handleLightweightExtensionFolderPick($event)">
+                        <div style="display:flex;justify-content:space-between;align-items:center;gap:8px;margin-top:8px;">
+                            <span>{{ lightweightExtensionTestStatus }}</span>
+                            <el-button size="small" type="success" @click="openLightweightExtensionFolderPicker">选择文件夹</el-button>
+                        </div>
+                    </div>
+                    <!-- lightWeightExtensionTest:end -->
                 </div>
             </div>
 
@@ -7955,6 +7973,7 @@ RNA-seq KEGG/fgsea 会优先使用物种级 KEGG gene set；没有物种级缓
                     working_dir: '',
                     output_dir: ''
                 });
+                const lightweightExtensionTestStatus = ref('拖入固定结构文件夹，系统将自动创建 Job、解析环境并开始 upload。');
                 const activeRnaseqScopeId = ref('');
                 const activeGenotypeScopeId = ref('');
 
@@ -8131,6 +8150,7 @@ RNA-seq KEGG/fgsea 会优先使用物种级 KEGG gene set；没有物种级缓
                         pop_snp_density: 'SNP密度分析',
                         rnaseq_de: 'RNA-seq DE',
                         combo: '组合图',
+                        gene_family_report: '基因家族结果报告',
                         reference_classification: '参考亚家族分类',
                         subfamily_confirm: '亚家族标签确认',
                         ppi_string_reference_confirmation: 'Confirm the STRING reference species for PPI',
@@ -9310,6 +9330,7 @@ RNA-seq KEGG/fgsea 会优先使用物种级 KEGG gene set；没有物种级缓
                     statusDialogVisible: false,
                     statusEvents: []
                 });
+                const extensionEnvOptions = ref(['test4py311', 'rnaseq', 'rnaR']);
                 const lightweightExtensions = ref([]);
                 const lightweightExtensionConversation = reactive({
                     cards: [],
@@ -10300,6 +10321,7 @@ compare: [[flesh, peel]]`;
                     { name: 'pep', label: '1️⃣2️⃣ PEP', status: 'pending', images: [], files: [], elapsed_time: 0 },
                     { name: 'expression', label: '1️⃣3️⃣ Expression (FPKM)', status: 'pending', images: [], files: [], elapsed_time: 0 },
                     { name: 'combo', label: '6️⃣ Combo', status: 'pending', images: [], files: [], elapsed_time: 0 },
+                    { name: 'gene_family_report', label: '📄 基因家族结果报告', status: 'pending', images: [], files: [], elapsed_time: 0 },
                     // RNA-seq stages
                     { name: 'rnaseq_qc', label: '1️⃣5️⃣ RNA-seq QC', status: 'pending', images: [], files: [], elapsed_time: 0 },
                     { name: 'rnaseq_align', label: '1️⃣6️⃣ RNA-seq Align', status: 'pending', images: [], files: [], elapsed_time: 0 },
@@ -15499,6 +15521,54 @@ compare: [[flesh, peel]]`;
                     return `⏸️ 批量执行已暂停：成功 ${completedCount}/${totalCount}，暂停 ${pausedCount}/${totalCount}。真实原因：${reasonText}。`;
                 };
 
+                const openGeneratedGeneFamilyReport = async () => {
+                    await refreshResults(jobInfo.job_id, 'gene_family_report');
+                    const reportStage = stages.find(stage => stage.name === 'gene_family_report');
+                    const reportFile = (reportStage?.files || []).find(file =>
+                        file.role === 'gene_family_report' || String(file.name || '').endsWith('.gene_family_report.md')
+                    );
+                    if (!reportFile) return false;
+                    rightPanelTab.value = 'results';
+                    selectFile('gene_family_report', reportFile);
+                    return true;
+                };
+
+                const submitGeneFamilyReportForBatch = async (stages, pausedCount) => {
+                    if (pausedCount > 0) return false;
+                    const plannedIds = (stages || []).map(stage => getStageId(stage)).filter(
+                        stageId => stageId && !['summary', 'gene_family_report', 'subfamily_confirm'].includes(stageId)
+                    );
+                    const geneFamilyStages = new Set([
+                        'identify', 'phylogeny', 'motif', 'domain', 'structure', 'promoter',
+                        'synteny', 'interspecies_synteny', 'karyotype', 'combo', 'ppi', 'expression', 'pep'
+                    ]);
+                    if (!plannedIds.some(stageId => geneFamilyStages.has(stageId))) return false;
+                    const stageStatuses = Object.fromEntries(plannedIds.map(stageId => {
+                        const status = stageExecutionStatus.value[stageId];
+                        return [stageId, status === 'completed' ? 'success' : status === 'failed' ? 'failed' : 'missing'];
+                    }));
+                    const batchParams = batchExecuteParams.value || {};
+                    const planParams = batchParams.plan?.params || {};
+                    const stageParams = (stages || []).find(stage => stage && typeof stage === 'object' && stage.params)?.params || {};
+                    try {
+                        await executeSingleStageWithParams('gene_family_report', {
+                            expectedStages: plannedIds,
+                            stageStatuses,
+                            reportAction: 'generate',
+                            reportUseModel: true,
+                            scopeId: batchParams.scopeId || planParams.scopeId || stageParams.scopeId || '',
+                            analysisRoute: batchParams.analysisRoute || planParams.analysisRoute || stageParams.analysisRoute || '',
+                            memberSetId: batchParams.memberSetId || planParams.memberSetId || stageParams.memberSetId || '',
+                            rnaseqScopeId: batchParams.rnaseqScopeId || planParams.rnaseqScopeId || stageParams.rnaseqScopeId || ''
+                        });
+                        addAgentMessage('assistant', '📄 基因家族结果报告已生成。');
+                        return true;
+                    } catch (err) {
+                        addAgentMessage('assistant', `⚠️ 基因家族结果报告生成失败：${err.message || err}`);
+                        return false;
+                    }
+                };
+
                 const confirmAndStartBatchExecute = async (event) => {
                     if (!event?.isTrusted) return;
                     batchExecutionUserConfirmed.value = true;
@@ -15616,6 +15686,7 @@ compare: [[flesh, peel]]`;
                         }).length;
                         const totalCount = stages.length;
 
+                        let continuationOpened = false;
                         if (failedCount > 0) {
                             addAgentMessage('assistant', `⚠️ 批量执行完成：成功 ${completedCount}/${totalCount}，失败 ${failedCount}/${totalCount}`);
                         } else if (pausedCount > 0) {
@@ -15623,13 +15694,16 @@ compare: [[flesh, peel]]`;
 	                        } else {
 	                            addAgentMessage('assistant', `🎉 批量执行完成！成功: ${completedCount}/${totalCount}`);
                                 if (isMethodologyDataPreparationContinuationCandidate(stages, completedCount, totalCount, failedCount, pausedCount)) {
-                                    await openMethodologyDataPreparationContinuation();
+                                    continuationOpened = await openMethodologyDataPreparationContinuation();
                                 } else if (isMethodologyPostConfirmContinuationCandidate(stages, completedCount, totalCount, failedCount, pausedCount)) {
-                                    await openMethodologyPostConfirmContinuation();
+                                    continuationOpened = await openMethodologyPostConfirmContinuation();
                                 } else if (isContextualPostConfirmContinuationCandidate(stages, completedCount, totalCount, failedCount, pausedCount)) {
-                                    await openContextualPostConfirmContinuation();
+                                    continuationOpened = await openContextualPostConfirmContinuation();
                                 }
 	                        }
+                            if (pausedCount === 0 && !continuationOpened) {
+                                await submitGeneFamilyReportForBatch(stages, pausedCount);
+                            }
                     }
                 };
 
@@ -18023,6 +18097,16 @@ compare: [[flesh, peel]]`;
                         requestBody.geoId = confirmParams.geoId || '';
                         requestBody.idTransform = confirmParams.idTransform || 'remove_T01';
                     }
+                    if (stageName === 'gene_family_report') {
+                        requestBody.expectedStages = activeStageParams.expectedStages || [];
+                        requestBody.stageStatuses = activeStageParams.stageStatuses || {};
+                        requestBody.reportAction = activeStageParams.reportAction || 'generate';
+                        requestBody.reportUseModel = activeStageParams.reportUseModel !== false;
+                        requestBody.scopeId = activeStageParams.scopeId || '';
+                        requestBody.analysisRoute = activeStageParams.analysisRoute || '';
+                        requestBody.memberSetId = activeStageParams.memberSetId || '';
+                        requestBody.rnaseqScopeId = activeStageParams.rnaseqScopeId || '';
+                    }
                     if (stageName === 'promoter_plot_revise') {
                         const elements = Array.isArray(confirmParams.promoterElements) ? confirmParams.promoterElements : [];
                         const categories = Array.isArray(confirmParams.promoterCategories) ? confirmParams.promoterCategories : [];
@@ -18113,6 +18197,9 @@ compare: [[flesh, peel]]`;
                         } else if (data.async === false) {
                             // 同步执行完成，直接返回结果
                             console.log(`[DEBUG-EXEC] ${stageName}: sync completed`);
+                            if (stageName === 'gene_family_report') {
+                                await openGeneratedGeneFamilyReport();
+                            }
                             return data;
                         } else {
                             throw createStructuredExecuteError(data);
@@ -19381,6 +19468,9 @@ compare: [[flesh, peel]]`;
                                 result: data
                             });
                             await handleStageCompleted(stageName, data, {});
+                            if (stageName === 'gene_family_report') {
+                                await openGeneratedGeneFamilyReport();
+                            }
 
                             // 更新阶段结果 - 如果有分stage的文件，则更新所有stage的文件
                             if (stageIdx >= 0 && data.result) {
@@ -19949,6 +20039,10 @@ compare: [[flesh, peel]]`;
                 };
 
                 const extensionFileKey = (file) => `${file?.name || ''}::${file?.size || 0}::${file?.lastModified || 0}`;
+                const extensionUploadPath = (file) => {
+                    const relative = String(file?.lightweightRelativePath || '').trim().replace(/^\/+/, '');
+                    return relative || String(file?.name || '').trim();
+                };
                 const appendExtensionFiles = (target, files) => {
                     const current = Array.isArray(target) ? target : [];
                     const merged = [...current];
@@ -19974,6 +20068,53 @@ compare: [[flesh, peel]]`;
                     if (selected.length) ElMessage.success(`已选择 ${selected.length} 个测试数据文件`);
                     event.target.value = '';
                 };
+                const handleLightweightExtensionFolderDrop = (event) => {
+                    const droppedFiles = Array.from(event?.dataTransfer?.files || []);
+                    if (!droppedFiles.length) {
+                        ElMessage.warning('未检测到可用文件');
+                        return;
+                    }
+
+                    const codeExtPattern = /\.(py|r|sh|bash|pl|js|ts|java|c|cc|cpp|go|rs|jl|rb|m|smk|nf|ipynb)$/i;
+                    const dataExtPattern = /\.(fastq|fq|gz|fa|fasta|fna|faa|pep|cds|gff|gff3|gtf|csv|tsv|txt|xlsx|xls|vcf|sam|bam)$/i;
+
+                    const normalized = droppedFiles.map((file) => {
+                        const relativePath = file.webkitRelativePath || file.name;
+                        if (!file.lightweightRelativePath) {
+                            Object.defineProperty(file, 'lightweightRelativePath', {
+                                value: relativePath,
+                                configurable: true
+                            });
+                        }
+                        return file;
+                    });
+
+                    const codeFiles = [];
+                    const dataFiles = [];
+                    normalized.forEach((file) => {
+                        const filePath = extensionUploadPath(file).toLowerCase();
+                        if (codeExtPattern.test(filePath)) {
+                            codeFiles.push(file);
+                        } else if (dataExtPattern.test(filePath)) {
+                            dataFiles.push(file);
+                        } else {
+                            codeFiles.push(file);
+                        }
+                    });
+
+                    extensionBuild.codeFiles = appendExtensionFiles(extensionBuild.codeFiles, codeFiles);
+                    extensionBuild.dataFiles = appendExtensionFiles(extensionBuild.dataFiles, dataFiles);
+
+                    if (!extensionBuild.displayName) {
+                        const firstPath = normalized[0]?.webkitRelativePath || normalized[0]?.name || '';
+                        const folder = firstPath.includes('/') ? firstPath.split('/')[0] : '';
+                        if (folder) {
+                            extensionBuild.displayName = folder.replace(/[^a-zA-Z0-9_-]+/g, '_');
+                        }
+                    }
+
+                    ElMessage.success(`已拖入 ${normalized.length} 个文件，已自动填充拓展包`);
+                };
                 const removeExtensionFile = (kind, file) => {
                     const key = extensionFileKey(file);
                     if (kind === 'code') {
@@ -20006,9 +20147,10 @@ compare: [[flesh, peel]]`;
                 };
                 const extensionApiUrl = (path) => {
                     const value = String(path || '');
-                    if (!value.startsWith('/api/extension-stages') && !value.startsWith('/api/extension-stage-runs')) return value;
+                    if (!value.startsWith('/api/extension-stages') && !value.startsWith('/api/extension-stage-runs') && !value.startsWith('/api/lightweight-extension-stages')) return value;
                     const host = window.location.hostname || '127.0.0.1';
-                    return `${window.location.protocol}//${host}:7880${value}`;
+                    const port = window.location.port;
+                    return port ? `${window.location.protocol}//${host}:${port}${value}` : `${window.location.protocol}//${host}${value}`;
                 };
                 const extensionApiFetch = async (path, options = {}) => fetch(extensionApiUrl(path), options);
                 const pushExtensionStatus = (message, type = 'info', showDialog = false) => {
@@ -20369,9 +20511,10 @@ compare: [[flesh, peel]]`;
                         form.append('extension_id', extensionId);
                         if (hasCommand) form.append('original_command', extensionBuild.command);
                         form.append('operation_token', token);
-                        extensionBuild.codeFiles.forEach(file => form.append('raw_code_files', file));
-                        extensionBuild.dataFiles.forEach(file => form.append('raw_data_files', file));
+                        extensionBuild.codeFiles.forEach(file => form.append('raw_code_files', file, extensionUploadPath(file)));
+                        extensionBuild.dataFiles.forEach(file => form.append('raw_data_files', file, extensionUploadPath(file)));
                         if (hasDataPaths) form.append('raw_data_paths', extensionBuild.dataPaths);
+                        form.append('runtime_env', extensionBuild.runtimeEnv);
                         const response = await extensionApiFetch('/api/lightweight-extension-stages/upload', { method: 'POST', body: form });
                         if (!response.ok) throw new Error(await extensionApiError(response, '原始包可复现检查提交失败'));
                         ElMessage.success('文件上传成功，已开始生成并验收');
@@ -20616,6 +20759,18 @@ compare: [[flesh, peel]]`;
                         console.warn('[lightweight extension] list failed', err);
                     }
                 };
+                const loadExtensionEnvOptions = async () => {
+                    try {
+                        const response = await extensionApiFetch('/api/extension-stages/envs');
+                        if (!response.ok) return;
+                        const data = await response.json();
+                        if (data.success && Array.isArray(data.envs) && data.envs.length) {
+                            extensionEnvOptions.value = data.envs;
+                        }
+                    } catch (err) {
+                        console.warn('[extension env options] load failed', err);
+                    }
+                };
                 const answerLightweightExtensionHelp = async (message) => {
                     const query = String(message || '');
                     const matched = (lightweightExtensions.value || []).find(item => query.includes(item.extension_id));
@@ -23857,6 +24012,7 @@ compare: [[flesh, peel]]`;
                     startRuntimeMonitor();
                     loadExtensionStages();
                     loadLightweightExtensionStages();
+                    loadExtensionEnvOptions();
                     resumeLightweightOperation();
                     loadExtensionRuns();
                     loadMethodologyJobs();
@@ -24189,6 +24345,7 @@ compare: [[flesh, peel]]`;
                     intentHistory,
                     uploadedFiles,
                     extensionBuild,
+                    extensionEnvOptions,
                     extensionStages,
                     extensionCompletedStages,
                     lightweightExtensionConversation,
@@ -24207,6 +24364,7 @@ compare: [[flesh, peel]]`;
                     extensionFileKey,
                     onExtensionCodeFilesChange,
                     onExtensionDataFilesChange,
+                    handleLightweightExtensionFolderDrop,
                     removeExtensionFile,
                     buildExtensionStage,
                     confirmLightweightCandidate,
@@ -24226,6 +24384,7 @@ compare: [[flesh, peel]]`;
                     validateExtensionStage,
                     publishExtensionStage,
                     loadExtensionStages,
+                    loadExtensionEnvOptions,
                     loadExtensionRuns,
                     runExtensionStage,
                     runExtensionStageFromIntent,
@@ -24345,6 +24504,24 @@ compare: [[flesh, peel]]`;
                     openPreview,
                     redrawImage,
                     createNewJob,
+                    lightweightExtensionTestStatus,
+                    handleLightweightExtensionFolderDrop: (event) => window.LightWeightExtensionTest && window.LightWeightExtensionTest.handleDrop({
+                        jobInfo,
+                        extensionBuild,
+                        createNewJob,
+                        appendExtensionFiles,
+                        buildLightweightExtensionStage,
+                        lightweightExtensionTestStatus,
+                    }, event),
+                    handleLightweightExtensionFolderPick: (event) => window.LightWeightExtensionTest && window.LightWeightExtensionTest.handlePick({
+                        jobInfo,
+                        extensionBuild,
+                        createNewJob,
+                        appendExtensionFiles,
+                        buildLightweightExtensionStage,
+                        lightweightExtensionTestStatus,
+                    }, event),
+                    openLightweightExtensionFolderPicker: () => window.LightWeightExtensionTest && window.LightWeightExtensionTest.openPicker(),
                     // Working DirectorySelect
                     existingJobs,
                     jobSelectorVisible,
@@ -24461,6 +24638,9 @@ compare: [[flesh, peel]]`;
         document.getElementById('app').style.display = 'block';
         console.log('App mounted successfully');
     </script>
+    <!-- lightWeightExtensionTest:wiring:start -->
+    <script src="/lightWeightExtensionTest.js"></script>
+    <!-- lightWeightExtensionTest:wiring:end -->
     <!-- 深度解读弹窗样式 -->
     <style>
         .interpretation-dialog .el-dialog__body { padding: 0; max-height: 70vh; overflow-y: auto; }

apps/web/vue3/kg_ml.html

@@ -0,0 +1,712 @@
+<!DOCTYPE html>
+<html lang="zh-CN">
+<head>
+    <meta charset="UTF-8">
+    <meta name="viewport" content="width=device-width, initial-scale=1.0">
+    <title>Knowledge Graph</title>
+    <link rel="stylesheet" href="https://cdn.jsdelivr.net/npm/element-plus@2.7.1/dist/index.css">
+    <script src="https://cdn.jsdelivr.net/npm/vue@3/dist/vue.global.prod.js"></script>
+    <script src="https://cdn.jsdelivr.net/npm/element-plus@2.7.1/dist/index.full.min.js"></script>
+    <script src="https://cdn.jsdelivr.net/npm/@element-plus/icons-vue"></script>
+    <script src="https://cdn.jsdelivr.net/npm/echarts@5.4.3/dist/echarts.min.js"></script>
+    <script src="/static/i18n.js"></script>
+    <style>
+        body { margin: 0; font-family: 'Helvetica Neue', Arial, sans-serif; background: #f5f7fa; }
+        .header { background: linear-gradient(135deg, #667eea 0%, #764ba2 100%); color: white; padding: 15px 20px; display: flex; align-items: center; justify-content: space-between; }
+        .header h1 { margin: 0; font-size: 22px; }
+        .header .subtitle { font-size: 12px; opacity: 0.8; margin-top: 4px; }
+        .main-container { display: flex; padding: 15px; gap: 15px; height: calc(100vh - 70px); }
+        .left-panel { width: 300px; display: flex; flex-direction: column; gap: 10px; }
+        .control-section { background: white; border-radius: 8px; padding: 15px; box-shadow: 0 2px 8px rgba(0,0,0,0.08); }
+        .control-section h4 { margin: 0 0 10px 0; color: #409EFF; font-size: 14px; border-bottom: 2px solid #409EFF; padding-bottom: 8px; }
+        .stat-item { display: flex; justify-content: space-between; padding: 8px 0; border-bottom: 1px solid #eee; }
+        .stat-item:last-child { border-bottom: none; }
+        .stat-label { color: #666; }
+        .stat-value { font-weight: bold; color: #409EFF; }
+        .filter-section { margin-top: 10px; }
+        .filter-tag { margin: 4px; cursor: pointer; }
+        .filter-tag.active { background: #409EFF; color: white; }
+        .main-panel { flex: 1; display: flex; flex-direction: column; background: white; border-radius: 8px; box-shadow: 0 2px 8px rgba(0,0,0,0.08); overflow: hidden; }
+        .toolbar { padding: 10px 15px; border-bottom: 1px solid #eee; display: flex; align-items: center; gap: 10px; }
+        #graph-container { flex: 1; position: relative; }
+        .right-panel { width: 350px; background: white; border-radius: 8px; box-shadow: 0 2px 8px rgba(0,0,0,0.08); padding: 15px; overflow-y: auto; }
+        .right-panel h4 { margin: 0 0 10px 0; color: #67C23A; border-bottom: 2px solid #67C23A; padding-bottom: 8px; }
+        .node-detail { padding: 10px; background: #f5f7fa; border-radius: 6px; margin-top: 10px; }
+        .node-detail .label { color: #666; font-size: 12px; margin-bottom: 4px; }
+        .node-detail .value { font-weight: 500; color: #333; }
+        .relation-list { margin-top: 10px; }
+        .relation-item { padding: 8px; background: #f0f9eb; border-radius: 4px; margin-bottom: 8px; font-size: 13px; }
+        .relation-item .source { color: #409EFF; font-weight: 500; }
+        .relation-item .rel { color: #67C23A; margin: 0 4px; }
+        .relation-item .target { color: #E6A23C; font-weight: 500; }
+        .search-box { margin-bottom: 10px; }
+        .legend { display: flex; flex-wrap: wrap; gap: 8px; margin-top: 10px; }
+        .legend-item { display: flex; align-items: center; gap: 4px; font-size: 12px; }
+        .legend-color { width: 12px; height: 12px; border-radius: 50%; }
+        .legend-symbol { width: 12px; height: 12px; border-radius: 2px; }
+        .legend-line { width: 20px; height: 2px; }
+        
+        /* PC语义分类样式 */
+        .pc-category-header {
+            font-size: 11px;
+            color: #666;
+            font-weight: bold;
+            margin: 8px 0 4px 0;
+            padding-left: 4px;
+            border-left: 3px solid #409EFF;
+        }
+    </style>
+</head>
+<body>
+    <div id="app">
+        <div class="header">
+            <div>
+                <h1>{{ t('kg-visualization') }}</h1>
+                <div class="subtitle">{{ t('kg-subtitle') }}</div>
+            </div>
+            <el-button type="primary" @click="goBack">
+                <el-icon><arrow-left /></el-icon> {{ t('back-to-main') }}
+            </el-button>
+        </div>
+        
+        <div class="main-container">
+            <!-- 左侧面板 -->
+            <div class="left-panel">
+                <div class="control-section">
+                    <h4>{{ t('kg-stats') }}</h4>
+                    <div class="stat-item">
+                        <span class="stat-label">{{ t('node-count') }}</span>
+                        <span class="stat-value">{{ stats.nodes }}</span>
+                    </div>
+                    <div class="stat-item">
+                        <span class="stat-label">{{ t('edge-count') }}</span>
+                        <span class="stat-value">{{ stats.edges }}</span>
+                    </div>
+                    <div class="stat-item">
+                        <span class="stat-label">{{ t('node-types') }}</span>
+                        <span class="stat-value">{{ stats.nodeTypes }}</span>
+                    </div>
+                    <div class="stat-item">
+                        <span class="stat-label">{{ t('edge-types') }}</span>
+                        <span class="stat-value">{{ stats.edgeTypes }}</span>
+                    </div>
+                </div>
+                
+                <div class="control-section">
+                    <h4>{{ t('node-search') }}</h4>
+                    <el-input
+                        v-model="searchQuery"
+                        :placeholder="t('enter-gene-name')"
+                        clearable
+                        @keyup.enter="searchNode"
+                    >
+                        <template #append>
+                            <el-button @click="searchNode">
+                                <el-icon><search /></el-icon>
+                            </el-button>
+                        </template>
+                    </el-input>
+                </div>
+                
+                <div class="control-section">
+                    <h4>{{ t('relation-filter-pc') }}</h4>
+                    <div class="filter-section">
+                        <div v-for="(types, category) in relationTypesByCategory" :key="category" style="margin-bottom: 10px;">
+                            <div style="font-size: 12px; color: #666; margin-bottom: 5px; font-weight: bold;">
+                                {{ category === 'activation' ? t('activation-induction') :
+                                   category === 'repression' ? t('repression-inhibition') :
+                                   category === 'regulation' ? t('regulation-control') :
+                                   category === 'expression' ? t('expression-detection') :
+                                   category === 'association' ? t('association-interaction') :
+                                   category === 'localization' ? t('localization') :
+                                   category === 'containment' ? t('containment-composition') :
+                                   category === 'participation' ? t('participation-function') : t('other-rel') }}
+                            </div>
+                            <el-tag 
+                                v-for="type in types" 
+                                :key="type"
+                                class="filter-tag"
+                                :class="{ active: selectedRelations.includes(type) }"
+                                @click="toggleRelationFilter(type)"
+                                size="small"
+                                :style="{ borderColor: relationEdgeStyles[category].color, color: selectedRelations.includes(type) ? '#fff' : relationEdgeStyles[category].color }"
+                            >
+                                {{ type }} ({{ relationCounts[type] }})
+                            </el-tag>
+                        </div>
+                    </div>
+                </div>
+                
+                <div class="control-section">
+                    <h4>{{ t('node-legend-pc') }}</h4>
+                    <div class="pc-category-header">Gene/Protein</div>
+                    <div class="legend">
+                        <div class="legend-item" v-for="type in ['gene', 'gene_id', 'gene_group', 'gene_family']" :key="type">
+                            <div class="legend-symbol" :style="{ background: nodeColors[type] }"></div>
+                            <span>{{ nodeTypeLabels[type] }}</span>
+                        </div>
+                    </div>
+                    <div class="pc-category-header">Phenotype</div>
+                    <div class="legend">
+                        <div class="legend-item" v-for="type in ['phenotype', 'attribute']" :key="type">
+                            <div class="legend-symbol" :style="{ background: nodeColors[type] }"></div>
+                            <span>{{ nodeTypeLabels[type] }}</span>
+                        </div>
+                    </div>
+                    <div class="pc-category-header">Cell/Organ/Organism</div>
+                    <div class="legend">
+                        <div class="legend-item" v-for="type in ['tissue', 'species']" :key="type">
+                            <div class="legend-symbol" :style="{ background: nodeColors[type] }"></div>
+                            <span>{{ nodeTypeLabels[type] }}</span>
+                        </div>
+                    </div>
+                    <div class="pc-category-header">Biological Process</div>
+                    <div class="legend">
+                        <div class="legend-item" v-for="type in ['process', 'time', 'condition', 'stress']" :key="type">
+                            <div class="legend-symbol" :style="{ background: nodeColors[type] }"></div>
+                            <span>{{ nodeTypeLabels[type] }}</span>
+                        </div>
+                    </div>
+                    <div class="pc-category-header">Genomic</div>
+                    <div class="legend">
+                        <div class="legend-item" v-for="type in ['location', 'sra_dataset', 'data_type', 'count']" :key="type">
+                            <div class="legend-symbol" :style="{ background: nodeColors[type] }"></div>
+                            <span>{{ nodeTypeLabels[type] }}</span>
+                        </div>
+                    </div>
+                    <h4 style="margin-top: 15px;">🔗 {{ t('edge-legend-pc') }}</h4>
+                    <div class="legend">
+                        <div class="legend-item" v-for="(style, cat) in relationEdgeStyles" :key="cat">
+                            <div class="legend-line" :style="{ background: style.color, height: style.width + 'px', borderTop: style.type === 'dashed' ? '2px dashed ' + style.color : 'none' }"></div>
+                            <span>{{ cat === 'activation' ? 'Activation/Induction' : 
+                                    cat === 'repression' ? 'Repression/Inhibition' :
+                                    cat === 'regulation' ? 'Regulation/Control' :
+                                    cat === 'expression' ? 'Expression/Detection' :
+                                    cat === 'association' ? 'Association/Interaction' :
+                                    cat === 'localization' ? 'Localization' :
+                                    cat === 'containment' ? 'Containment/Composition' :
+                                    cat === 'participation' ? 'Participation/Function' : 'Other' }}</span>
+                        </div>
+                    </div>
+                </div>
+            </div>
+            
+            <!-- 主面板 -->
+            <div class="main-panel">
+                <div class="toolbar">
+                    <el-button-group>
+                        <el-button @click="fitGraph">
+                            <el-icon><full-screen /></el-icon> {{ t('fit-view') }}
+                        </el-button>
+                        <el-button @click="resetFilters">
+                            <el-icon><refresh /></el-icon> {{ t('reset-filters') }}
+                        </el-button>
+                    </el-button-group>
+                    <el-divider direction="vertical"></el-divider>
+                    <span style="color: #666; font-size: 13px;">
+                        {{ t('click-drag-hint') }}
+                    </span>
+                </div>
+                <div id="graph-container"></div>
+            </div>
+            
+            <!-- 右侧面板 -->
+            <div class="right-panel">
+                <h4>{{ t('node-detail') }}</h4>
+                <div v-if="selectedNode" class="node-detail">
+                    <div style="margin-bottom: 15px;">
+                        <div class="label">{{ t('node-name') }}</div>
+                        <div class="value" style="font-size: 18px; color: #409EFF;">{{ selectedNode.name }}</div>
+                    </div>
+                    <div style="margin-bottom: 15px;">
+                        <div class="label">{{ t('node-type-pc') }}</div>
+                        <div class="value">
+                            <el-tag :type="getNodeTypeTag(selectedNode.category)" :style="{ backgroundColor: nodeColors[selectedNode.category] + '20', borderColor: nodeColors[selectedNode.category], color: nodeColors[selectedNode.category] }">
+                                {{ nodeTypeLabels[selectedNode.category] || selectedNode.category }}
+                            </el-tag>
+                        </div>
+                    </div>
+                    <div>
+                        <div class="label">{{ t('connection-count') }}</div>
+                        <div class="value">{{ selectedNode.degree }}{{ t('num-connections') }}</div>
+                    </div>
+                </div>
+                <div v-else style="color: #999; text-align: center; padding: 40px;">
+                    {{ t('click-node-detail') }}
+                </div>
+                
+                <template v-if="selectedNodeRelations.length > 0">
+                    <h4 style="margin-top: 20px;">🔗 {{ t('related-relations') }} ({{ selectedNodeRelations.length }})</h4>
+                    <div class="relation-list">
+                        <div 
+                            v-for="rel in selectedNodeRelations.slice(0, 10)" 
+                            :key="rel.id"
+                            class="relation-item"
+                            :style="{ borderLeft: '3px solid ' + relationEdgeStyles[rel.category || relationCategories[rel.name] || 'other'].color }"
+                        >
+                            <span class="source">{{ rel.source }}</span>
+                            <span class="rel" :style="{ color: relationEdgeStyles[rel.category || relationCategories[rel.name] || 'other'].color }">--{{ rel.name }}--></span>
+                            <span class="target">{{ rel.target }}</span>
+                        </div>
+                        <div v-if="selectedNodeRelations.length > 10" style="text-align: center; color: #999; padding: 10px;">
+                            {{ tp('more-relations', { count: selectedNodeRelations.length - 10 }) }}
+                        </div>
+                    </div>
+                </template>
+            </div>
+        </div>
+    </div>
+
+    <script>
+        const { createApp, ref, onMounted, computed } = Vue;
+        const { ElMessage } = ElementPlus;
+
+        createApp({
+            setup() {
+                const t = window.t || ((key) => key);
+                const tp = window.tp || ((key, params) => {
+                    let s = t(key);
+                    if (params) Object.keys(params).forEach(k => { s = s.replace('{' + k + '}', params[k]); });
+                    return s;
+                });
+                document.title = t('kg-title');
+                const graphData = ref({ nodes: [], edges: [] });
+                const chart = ref(null);
+                const searchQuery = ref('');
+                const selectedNode = ref(null);
+                const selectedRelations = ref([]);
+                
+                // PC (Pathway Commons) 语义分类颜色映射
+                const nodeColors = {
+                    // Gene/Protein - 蓝色
+                    'gene': '#5DADE2',
+                    'gene_id': '#85C1E9',
+                    'gene_group': '#AED6F1',
+                    'gene_family': '#5DADE2',
+                    
+                    // Phenotype - 绿色
+                    'phenotype': '#58D68D',
+                    'attribute': '#82E0AA',
+                    
+                    // Cell/Organ/Organism - 粉色
+                    'tissue': '#F5B7B1',
+                    'species': '#F1948A',
+                    
+                    // Process - 紫色菱形
+                    'process': '#AF7AC5',
+                    'time': '#C39BD3',
+                    'condition': '#D7BDE2',
+                    'stress': '#E8DAEF',
+                    
+                    // Location - 浅绿色
+                    'location': '#82E0AA',
+                    
+                    // Data - 橙色（更显眼）
+                    'sra_dataset': '#E67E22',
+                    'data_type': '#F39C12',
+                    'count': '#EAEDED',
+                    
+                    'unknown': '#95A5A6'
+                ,
+            'article': '#2ecc71',
+            'paper_title': '#9b59b6'
+        };
+                
+                // PC 语义分类显示名称
+                const nodeTypeLabels = {
+                    'gene': 'Gene/Protein',
+                    'gene_id': 'Gene Identifier',
+                    'gene_group': 'Gene Group',
+                    'gene_family': 'Gene Family',
+                    'phenotype': 'Phenotype',
+                    'attribute': 'Attribute',
+                    'tissue': 'Cell/Organ/Tissue',
+                    'species': 'Organism',
+                    'process': 'Biological Process',
+                    'time': 'Time/Stage',
+                    'condition': 'Condition',
+                    'stress': 'Stress',
+                    'location': 'Genomic Location',
+                    'sra_dataset': 'SRA Dataset',
+                    'geo_dataset': 'GEO Dataset',
+                    'data_type': 'Data Type',
+                    'count': 'Count/Statistic',
+                    'article': 'Article (PMID)',
+                    'paper_title': 'Paper Title',
+                    'unknown': 'Other'
+                };
+                
+                // PC 关系类型分类
+                const relationCategories = {
+                    // Activation/Induction - 绿色箭头
+                    'induces': 'activation',
+                    'promote': 'activation',
+                    'overexpression_increases': 'activation',
+                    'enhanced_tolerance_to': 'activation',
+                    
+                    // Repression/Inhibition - 红色T型
+                    'down_regulated_during': 'repression',
+                    'down_regulation_improves': 'repression',
+                    
+                    // Regulation - 蓝色
+                    'regulate': 'regulation',
+                    'regulates': 'regulation',
+                    'regulating': 'regulation',
+                    
+                    // Expression - 橙色
+                    'expressed_in': 'expression',
+                    'expressed_during': 'expression',
+                    'was_expressed_during': 'expression',
+                    'expression_reached_peak_at': 'expression',
+                    
+                    // Association - 紫色
+                    'collinear_with': 'association',
+                    'was_highly_orthologous_to': 'association',
+                    'related_to': 'association',
+                    'affect': 'association',
+                    
+                    // Localization - 深绿
+                    'located_on': 'localization',
+                    'identified_in': 'localization',
+                    
+                    // Containment - 青色
+                    'contain': 'containment',
+                    'had': 'containment',
+                    'has_gene_id': 'containment',
+                    'has_data_type': 'containment',
+                    
+                    // Participation - 深红
+                    'involved_in': 'participation',
+                    'participate_in': 'participation',
+                    
+                    'default': 'other'
+                };
+                
+                const relationEdgeStyles = {
+                    'activation': { color: '#27AE60', width: 2, type: 'solid' },
+                    'repression': { color: '#E74C3C', width: 2, type: 'solid' },
+                    'regulation': { color: '#3498DB', width: 2, type: 'solid' },
+                    'expression': { color: '#E67E22', width: 1.5, type: 'solid' },
+                    'association': { color: '#9B59B6', width: 1.5, type: 'dashed' },
+                    'localization': { color: '#16A085', width: 1.5, type: 'solid' },
+                    'containment': { color: '#1ABC9C', width: 1.5, type: 'solid' },
+                    'participation': { color: '#C0392B', width: 2, type: 'solid' },
+                    'other': { color: '#95A5A6', width: 1, type: 'solid' }
+                };
+                
+                const stats = computed(() => ({
+                    nodes: graphData.value.nodes.length,
+                    edges: graphData.value.edges.length,
+                    nodeTypes: Object.keys(nodeColors).length,
+                    edgeTypes: relationTypes.value.length
+                }));
+                
+                // 按 PC 语义分类的关系类型
+                const relationTypesByCategory = computed(() => {
+                    const cats = {};
+                    graphData.value.edges.forEach(e => {
+                        const cat = relationCategories[e.relation] || 'other';
+                        if (!cats[cat]) cats[cat] = [];
+                        cats[cat].push(e.relation);
+                    });
+                    // 去重并排序
+                    for (const cat in cats) {
+                        cats[cat] = [...new Set(cats[cat])].sort();
+                    }
+                    return cats;
+                });
+                
+                const relationTypes = computed(() => {
+                    const types = new Set();
+                    graphData.value.edges.forEach(e => types.add(e.relation));
+                    return Array.from(types).sort();
+                });
+                
+                const relationCounts = computed(() => {
+                    const counts = {};
+                    graphData.value.edges.forEach(e => {
+                        counts[e.relation] = (counts[e.relation] || 0) + 1;
+                    });
+                    return counts;
+                });
+                
+                const selectedNodeRelations = computed(() => {
+                    if (!selectedNode.value) return [];
+                    return graphData.value.edges.filter(e => 
+                        e.source === selectedNode.value.id || e.target === selectedNode.value.id
+                    );
+                });
+                
+                const loadData = async () => {
+                    try {
+                        // 优先加载包含多组学数据的知识图谱
+                        let response = await fetch('knowledge_graph_with_omics.json');
+                        if (!response.ok) {
+                            // 回退到标准知识图谱
+                            response = await fetch('knowledge_graph.json');
+                        }
+                        if (!response.ok) {
+                            throw new Error(`HTTP ${response.status}: ${response.statusText}`);
+                        }
+                        const data = await response.json();
+                        
+                        // 验证数据结构
+                        if (!data.nodes || !Array.isArray(data.nodes)) {
+                            throw new Error(t('err-data-format'));
+                        }
+                        if (!data.edges || !Array.isArray(data.edges)) {
+                            console.warn(t('warn-edges-undefined'));
+                            data.edges = [];
+                        }
+                        
+                        // 计算节点度数
+                        const nodeDegrees = {};
+                        data.edges.forEach(e => {
+                            nodeDegrees[e.source] = (nodeDegrees[e.source] || 0) + 1;
+                            nodeDegrees[e.target] = (nodeDegrees[e.target] || 0) + 1;
+                        });
+                        
+                        // 根据 PC 语义分类设置节点形状
+                        const getNodeSymbol = (type) => {
+                            const symbolMap = {
+                                'gene': 'circle',
+                                'gene_id': 'circle',
+                                'gene_group': 'circle',
+                                'gene_family': 'circle',
+                                'phenotype': 'rect',
+                                'attribute': 'rect',
+                                'tissue': 'roundRect',
+                                'species': 'roundRect',
+                                'process': 'diamond',
+                                'time': 'diamond',
+                                'condition': 'diamond',
+                                'stress': 'diamond',
+                                'location': 'triangle',
+                                'sra_dataset': 'pin',
+                                'data_type': 'pin',
+                                'count': 'pin'
+                            };
+                            return symbolMap[type] || 'circle';
+                        };
+                        
+                        graphData.value = {
+                            nodes: data.nodes.map(n => ({
+                                id: n.id,
+                                name: n.id,
+                                category: n.type || 'unknown',
+                                degree: nodeDegrees[n.id] || 0,
+                                symbol: getNodeSymbol(n.type),
+                                symbolSize: (n.type === 'gene' ? 20 : n.type === 'sra_dataset' ? 18 : 12) + Math.min((nodeDegrees[n.id] || 0) * 0.3, 15),
+                                itemStyle: { 
+                                    color: nodeColors[n.type] || nodeColors.unknown,
+                                    borderColor: '#fff',
+                                    borderWidth: 1
+                                }
+                            })),
+                            edges: data.edges.map((e, i) => {
+                                const cat = relationCategories[e.relation] || 'other';
+                                const style = relationEdgeStyles[cat];
+                                return {
+                                    id: i,
+                                    source: e.source,
+                                    target: e.target,
+                                    name: e.relation,
+                                    relation: e.relation,
+                                    category: cat,
+                                    lineStyle: {
+                                        color: style.color,
+                                        width: style.width,
+                                        type: style.type,
+                                        curveness: 0.1
+                                    }
+                                };
+                            })
+                        };
+                        
+                        initChart();
+                        ElMessage.success(t('msg-kg-loaded'));
+                    } catch (error) {
+                        ElMessage.error(t('msg-load-data-failed') + error.message);
+                    }
+                };
+                
+                const initChart = () => {
+                    const container = document.getElementById('graph-container');
+                    chart.value = echarts.init(container);
+                    
+                    const option = {
+                        tooltip: {
+                            formatter: params => {
+                                if (params.dataType === 'node') {
+                                    const pcType = nodeTypeLabels[params.data.category] || params.data.category;
+                                    return `<b>${params.name}</b><br/>PC: ${pcType}<br/>${t('connection-count').replace('Connections', 'Deg')}: ${params.data.degree}`;
+                                }
+                                const cat = params.data.category || relationCategories[params.name] || 'other';
+                                const catName = cat === 'activation' ? 'Activation' : 
+                                               cat === 'repression' ? 'Repression' :
+                                               cat === 'regulation' ? 'Regulation' :
+                                               cat === 'expression' ? 'Expression' :
+                                               cat === 'association' ? 'Association' :
+                                               cat === 'localization' ? 'Localization' :
+                                               cat === 'containment' ? 'Containment' :
+                                               cat === 'participation' ? 'Participation' : 'Other';
+                                return `<b>${params.data.source}</b> <span style="color:${relationEdgeStyles[cat].color}">--${params.name}--></span> <b>${params.data.target}</b><br/><small>PC: ${catName}</small>`;
+                            }
+                        },
+                        series: [{
+                            type: 'graph',
+                            layout: 'force',
+                            data: graphData.value.nodes,
+                            links: graphData.value.edges,
+                            roam: true,
+                            draggable: true,
+                            label: {
+                                show: true,
+                                position: 'right',
+                                formatter: '{b}'
+                            },
+                            force: {
+                                repulsion: 300,
+                                gravity: 0.1,
+                                edgeLength: 100,
+                                layoutAnimation: true
+                            },
+                            emphasis: {
+                                focus: 'adjacency',
+                                lineStyle: { width: 4 }
+                            }
+                        }]
+                    };
+                    
+                    chart.value.setOption(option);
+                    
+                    chart.value.on('click', params => {
+                        if (params.dataType === 'node') {
+                            selectedNode.value = params.data;
+                        } else {
+                            selectedNode.value = null;
+                        }
+                    });
+                    
+                    window.addEventListener('resize', () => chart.value.resize());
+                };
+                
+                const searchNode = () => {
+                    if (!searchQuery.value || !chart.value) return;
+                    
+                    const node = graphData.value.nodes.find(n => 
+                        n.id.toLowerCase().includes(searchQuery.value.toLowerCase())
+                    );
+                    
+                    if (node) {
+                        chart.value.dispatchAction({
+                            type: 'highlight',
+                            seriesIndex: 0,
+                            name: node.id
+                        });
+                        selectedNode.value = node;
+                        ElMessage.success(t('msg-node-found') + node.name);
+                    } else {
+                        ElMessage.warning(t('msg-node-not-found'));
+                    }
+                };
+                
+                const toggleRelationFilter = (type) => {
+                    const index = selectedRelations.value.indexOf(type);
+                    if (index > -1) {
+                        selectedRelations.value.splice(index, 1);
+                    } else {
+                        selectedRelations.value.push(type);
+                    }
+                    
+                    if (selectedRelations.value.length === 0) {
+                        chart.value.setOption({
+                            series: [{ data: graphData.value.nodes, links: graphData.value.edges }]
+                        });
+                    } else {
+                        const filteredEdges = graphData.value.edges.filter(e => 
+                            selectedRelations.value.includes(e.relation)
+                        );
+                        const connectedNodes = new Set();
+                        filteredEdges.forEach(e => {
+                            connectedNodes.add(e.source);
+                            connectedNodes.add(e.target);
+                        });
+                        const filteredNodes = graphData.value.nodes.filter(n => 
+                            connectedNodes.has(n.id)
+                        );
+                        
+                        chart.value.setOption({
+                            series: [{ data: filteredNodes, links: filteredEdges }]
+                        });
+                    }
+                };
+                
+                const fitGraph = () => {
+                    if (chart.value) {
+                        chart.value.dispatchAction({ type: 'restore' });
+                    }
+                };
+                
+                const resetFilters = () => {
+                    selectedRelations.value = [];
+                    selectedNode.value = null;
+                    searchQuery.value = '';
+                    if (chart.value) {
+                        chart.value.setOption({
+                            series: [{ data: graphData.value.nodes, links: graphData.value.edges }]
+                        });
+                        chart.value.dispatchAction({ type: 'restore' });
+                    }
+                };
+                
+                const getNodeTypeTag = (type) => {
+                    const typeMap = { 'gene': '', 'gene_id': 'success', 'tissue': 'warning', 'location': 'info' };
+                    return typeMap[type] || '';
+                };
+                
+                const goBack = () => {
+                    window.location.href = '/';
+                };
+                
+                onMounted(() => {
+                    loadData().then(() => {
+                        // 检查URL参数，如果有highlight则自动搜索高亮
+                        const urlParams = new URLSearchParams(window.location.search);
+                        const highlightEntity = urlParams.get('highlight');
+                        if (highlightEntity) {
+                            searchQuery.value = highlightEntity;
+                            setTimeout(() => {
+                                searchNode();
+                            }, 500); // 等待图表初始化完成
+                        }
+                    });
+                });
+                
+                return {
+                    t,
+                    tp,
+                    stats,
+                    searchQuery,
+                    selectedNode,
+                    selectedRelations,
+                    relationTypes,
+                    relationTypesByCategory,
+                    relationCounts,
+                    relationEdgeStyles,
+                    relationCategories,
+                    nodeTypeLabels,
+                    selectedNodeRelations,
+                    nodeColors,
+                    searchNode,
+                    toggleRelationFilter,
+                    fitGraph,
+                    resetFilters,
+                    getNodeTypeTag,
+                    goBack
+                };
+            }
+        }).use(ElementPlus).mount('#app');
+    </script>
+</body>
+</html>

apps/web/vue3/lightWeightExtensionTest.js

@@ -0,0 +1,304 @@
+(function () {
+    const PY_EXTENSIONS = new Set(['.py']);
+    const R_EXTENSIONS = new Set(['.r']);
+
+    function normalizedRelativePath(file) {
+        const relative = String(file.webkitRelativePath || file.relativePath || file.name || '').replace(/^\/+/, '');
+        return relative.split('\\').join('/');
+    }
+
+    function parsePythonImports(content) {
+        const imports = new Set();
+        const lines = String(content || '').split(/\r?\n/);
+        for (const line of lines) {
+            let match = line.match(/^\s*import\s+([A-Za-z0-9_.,\s]+)/);
+            if (match) {
+                match[1].split(',').forEach((item) => {
+                    const token = item.trim().split(/\s+as\s+/i)[0].split('.')[0].trim();
+                    if (token) imports.add(token);
+                });
+            }
+            match = line.match(/^\s*from\s+([A-Za-z0-9_.]+)\s+import\s+/);
+            if (match) {
+                const token = match[1].split('.')[0].trim();
+                if (token) imports.add(token);
+            }
+        }
+        return Array.from(imports);
+    }
+
+    function parseRPackages(content) {
+        const packages = new Set();
+        const pattern = /(?:library|require)\(\s*["']?([A-Za-z0-9_.-]+)["']?\s*\)/g;
+        let match;
+        const text = String(content || '');
+        while ((match = pattern.exec(text)) !== null) {
+            if (match[1]) packages.add(match[1]);
+        }
+        return Array.from(packages);
+    }
+
+    function commonPathPrefix(files) {
+        const paths = files.map((f) => normalizedRelativePath(f).split('/')).filter((parts) => parts.length > 0);
+        if (!paths.length) return '';
+        const first = paths[0];
+        let prefixLen = 0;
+        for (let i = 0; i < first.length; i++) {
+            if (!paths.every((parts) => parts[i] === first[i])) break;
+            prefixLen = i + 1;
+        }
+        return first.slice(0, prefixLen).join('/');
+    }
+
+    function sanitizeDisplayName(raw) {
+        return String(raw || '').replace(/[^A-Za-z0-9_-]+/g, '_').replace(/^_+|_+$/g, '') || 'lightweight_extension';
+    }
+
+    function extractRootName(files) {
+        const prefix = commonPathPrefix(files);
+        const base = (prefix && prefix !== 'raw_code' && prefix !== 'raw_test_data')
+            ? sanitizeDisplayName(prefix.split('/')[0])
+            : 'lightweight_extension_drop';
+        const ts = Date.now();
+        return base + '_' + ts;
+    }
+
+    function splitFixedStructure(files) {
+        const codeFiles = [];
+        const dataFiles = [];
+        let hasRunScript = false;
+        let prefix = commonPathPrefix(files);
+        if (prefix && !prefix.endsWith('/')) prefix += '/';
+
+        const sampled = [];
+        for (const file of files) {
+            let relativePath = normalizedRelativePath(file);
+            if (prefix && relativePath.startsWith(prefix)) {
+                relativePath = relativePath.slice(prefix.length);
+            }
+            if (sampled.length < 10) sampled.push(relativePath);
+            if (relativePath.startsWith('raw_code/')) {
+                file.lightweightRelativePath = relativePath.slice('raw_code/'.length) || file.name;
+                codeFiles.push(file);
+                if (relativePath === 'raw_code/run.sh') {
+                    hasRunScript = true;
+                }
+            } else if (relativePath.startsWith('raw_test_data/')) {
+                file.lightweightRelativePath = relativePath.slice('raw_test_data/'.length) || file.name;
+                dataFiles.push(file);
+            }
+        }
+        if (!codeFiles.length) {
+            const preview = sampled.length ? sampled.slice(0, 10).join(', ') : '(无)';
+            throw new Error('缺少 raw_code/ 目录（检测到的路径样例: ' + preview + '）');
+        }
+        if (!dataFiles.length) {
+            throw new Error('缺少 raw_test_data/ 目录');
+        }
+        if (!hasRunScript) {
+            throw new Error('缺少 raw_code/run.sh');
+        }
+        return { codeFiles, dataFiles };
+    }
+
+    async function extractDependencies(files) {
+        const pythonImports = new Set();
+        const rPackages = new Set();
+        for (const file of files) {
+            const path = normalizedRelativePath(file);
+            const lower = path.toLowerCase();
+            const extension = lower.includes('.') ? lower.slice(lower.lastIndexOf('.')) : '';
+            if (!PY_EXTENSIONS.has(extension) && !R_EXTENSIONS.has(extension)) {
+                continue;
+            }
+            const content = await file.text();
+            if (PY_EXTENSIONS.has(extension)) {
+                parsePythonImports(content).forEach((name) => pythonImports.add(name));
+            } else {
+                parseRPackages(content).forEach((name) => rPackages.add(name));
+            }
+        }
+        return {
+            python_imports: Array.from(pythonImports),
+            r_packages: Array.from(rPackages),
+        };
+    }
+
+    async function encodeBase64(file) {
+        const buffer = await file.arrayBuffer();
+        const bytes = new Uint8Array(buffer);
+        let binary = '';
+        const chunkSize = 0x8000;
+        for (let index = 0; index < bytes.length; index += chunkSize) {
+            binary += String.fromCharCode(...bytes.subarray(index, index + chunkSize));
+        }
+        return btoa(binary);
+    }
+
+    async function importFolderToJob(hooks, rootName, files) {
+        const payload = {
+            job_id: String(hooks.jobInfo.job_id || '').trim(),
+            root_name: rootName,
+            files: [],
+            relative_paths: [],
+        };
+        for (const file of files) {
+            payload.files.push(await encodeBase64(file));
+            payload.relative_paths.push(normalizedRelativePath(file));
+        }
+        const response = await fetch('/api/job/import-folder', {
+            method: 'POST',
+            headers: { 'Content-Type': 'application/json' },
+            body: JSON.stringify(payload),
+        });
+        const data = await response.json();
+        if (!response.ok || !data.success) {
+            throw new Error(data.detail || data.error || data.message || 'Job 导入失败');
+        }
+        return data;
+    }
+
+    async function resolveRuntimeEnv(files) {
+        const dependencies = await extractDependencies(files);
+        console.log('[lightWeightExtensionTest] 提取的依赖:', JSON.stringify(dependencies), '| 扫描文件数:', files.length);
+        try {
+            const response = await fetch('/api/extension-stages/envs/resolve', {
+                method: 'POST',
+                headers: { 'Content-Type': 'application/json' },
+                body: JSON.stringify(dependencies),
+            });
+            const data = await response.json();
+            console.log('[lightWeightExtensionTest] 后端 env/resolve 原始响应:', response.status, JSON.stringify(data).slice(0, 500));
+            if (!response.ok || !data.success) {
+                throw new Error(data.detail || data.error || data.message || '运行环境解析失败');
+            }
+            return data;
+        } catch (e) {
+            console.error('[lightWeightExtensionTest] env/resolve 调用失败:', e);
+            throw e;
+        }
+    }
+
+    function notifyError(message) {
+        if (window.ElementPlus && window.ElementPlus.ElMessage) {
+            window.ElementPlus.ElMessage.error(message);
+        }
+    }
+
+    async function ensureJob(hooks) {
+        if (String(hooks.jobInfo.job_id || '').trim() && String(hooks.jobInfo.job_id) !== 'default_job') {
+            return;
+        }
+        await hooks.createNewJob();
+        if (!String(hooks.jobInfo.job_id || '').trim() || String(hooks.jobInfo.job_id) === 'default_job') {
+            throw new Error('自动创建 Job 失败');
+        }
+    }
+
+    async function processFolderFiles(hooks, inputFiles) {
+        const files = Array.from(inputFiles || []).filter((file) => normalizedRelativePath(file));
+        if (!files.length) {
+            throw new Error('未读取到任何文件');
+        }
+        hooks.lightweightExtensionTestStatus.value = '正在校验目录结构...';
+        const rootName = extractRootName(files);
+        const fixed = splitFixedStructure(files);
+        hooks.lightweightExtensionTestStatus.value = '正在确保 Job 可用...';
+        await ensureJob(hooks);
+        hooks.lightweightExtensionTestStatus.value = '正在解析运行环境...';
+        const envResult = await resolveRuntimeEnv(files);
+        hooks.extensionBuild.runtimeEnv = envResult.runtime_env || 'test4py311';
+        console.log(
+            '[lightWeightExtensionTest] ✓ 选定环境: ' + hooks.extensionBuild.runtimeEnv,
+            '| 原因: ' + (envResult.reason || 'unknown'),
+            '| 兜底: ' + (envResult.fallback ? '是' : '否'),
+            envResult.matches ? '| 候选: ' + JSON.stringify(envResult.matches.map(m => m.env + '(py:' + (m.python_hits || []).join(',') + ' r:' + (m.r_hits || []).join(',') + ')')) : '',
+        );
+        hooks.lightweightExtensionTestStatus.value = '正在镜像导入当前 Job...';
+        await importFolderToJob(hooks, rootName, files);
+        hooks.extensionBuild.displayName = rootName;
+        hooks.extensionBuild.command = 'bash run.sh';
+        hooks.extensionBuild.codeFiles = hooks.appendExtensionFiles([], fixed.codeFiles);
+        hooks.extensionBuild.dataFiles = hooks.appendExtensionFiles([], fixed.dataFiles);
+        hooks.extensionBuild.dataPaths = '';
+        hooks.lightweightExtensionTestStatus.value = `运行环境 ${hooks.extensionBuild.runtimeEnv} 已确定，正在触发 upload...`;
+        await hooks.buildLightweightExtensionStage();
+        hooks.lightweightExtensionTestStatus.value = '已开始 upload，系统正在自动轮询并等待候选卡。';
+    }
+
+    async function readEntry(entry, basePath) {
+        if (entry.isFile) {
+            return new Promise((resolve) => {
+                entry.file((file) => {
+                    Object.defineProperty(file, 'webkitRelativePath', {
+                        value: basePath ? basePath + '/' + entry.name : entry.name,
+                        writable: false,
+                    });
+                    resolve([file]);
+                });
+            });
+        }
+        if (entry.isDirectory) {
+            const reader = entry.createReader();
+            const entries = await new Promise((resolve) => {
+                reader.readEntries(resolve);
+            });
+            const subPath = basePath ? basePath + '/' + entry.name : entry.name;
+            const nested = await Promise.all(entries.map((e) => readEntry(e, subPath)));
+            return nested.flat();
+        }
+        return [];
+    }
+
+    async function readDroppedFiles(dataTransfer) {
+        if (!dataTransfer || !dataTransfer.items) {
+            return Array.from(dataTransfer?.files || []);
+        }
+        const items = Array.from(dataTransfer.items);
+        const entries = items
+            .map((item) => item.webkitGetAsEntry && item.webkitGetAsEntry())
+            .filter(Boolean);
+        if (!entries.length) {
+            return Array.from(dataTransfer.files || []);
+        }
+        const results = await Promise.all(entries.map((entry) => readEntry(entry, '')));
+        return results.flat();
+    }
+
+    function mountApi() {
+        window.LightWeightExtensionTest = {
+            async handleDrop(hooks, event) {
+                try {
+                    const files = await readDroppedFiles(event?.dataTransfer);
+                    await processFolderFiles(hooks, files);
+                } catch (error) {
+                    const message = `失败：${error.message || '目录处理失败'}`;
+                    hooks.lightweightExtensionTestStatus.value = message;
+                    notifyError(message);
+                }
+            },
+            async handlePick(hooks, event) {
+                const files = event?.target?.files || [];
+                try {
+                    await processFolderFiles(hooks, files);
+                } catch (error) {
+                    const message = `失败：${error.message || '目录处理失败'}`;
+                    hooks.lightweightExtensionTestStatus.value = message;
+                    notifyError(message);
+                } finally {
+                    if (event?.target) {
+                        event.target.value = '';
+                    }
+                }
+            },
+            openPicker() {
+                const input = document.getElementById('lightweight-extension-test-folder-input');
+                if (input) {
+                    input.click();
+                }
+            },
+        };
+    }
+
+    mountApi();
+})();

apps/web/vue3/static/cdn/en.js

@@ -0,0 +1,165 @@
+// global wrapper
+/*! Element Plus v2.3.14 */
+
+window.en = {
+  name: "en",
+  el: {
+    colorpicker: {
+      confirm: "OK",
+      clear: "Clear",
+      defaultLabel: "color picker",
+      description: "current color is {color}. press enter to select a new color."
+    },
+    datepicker: {
+      now: "Now",
+      today: "Today",
+      cancel: "Cancel",
+      clear: "Clear",
+      confirm: "OK",
+      dateTablePrompt: "Use the arrow keys and enter to select the day of the month",
+      monthTablePrompt: "Use the arrow keys and enter to select the month",
+      yearTablePrompt: "Use the arrow keys and enter to select the year",
+      selectedDate: "Selected date",
+      selectDate: "Select date",
+      selectTime: "Select time",
+      startDate: "Start Date",
+      startTime: "Start Time",
+      endDate: "End Date",
+      endTime: "End Time",
+      prevYear: "Previous Year",
+      nextYear: "Next Year",
+      prevMonth: "Previous Month",
+      nextMonth: "Next Month",
+      year: "",
+      month1: "January",
+      month2: "February",
+      month3: "March",
+      month4: "April",
+      month5: "May",
+      month6: "June",
+      month7: "July",
+      month8: "August",
+      month9: "September",
+      month10: "October",
+      month11: "November",
+      month12: "December",
+      week: "week",
+      weeks: {
+        sun: "Sun",
+        mon: "Mon",
+        tue: "Tue",
+        wed: "Wed",
+        thu: "Thu",
+        fri: "Fri",
+        sat: "Sat"
+      },
+      weeksFull: {
+        sun: "Sunday",
+        mon: "Monday",
+        tue: "Tuesday",
+        wed: "Wednesday",
+        thu: "Thursday",
+        fri: "Friday",
+        sat: "Saturday"
+      },
+      months: {
+        jan: "Jan",
+        feb: "Feb",
+        mar: "Mar",
+        apr: "Apr",
+        may: "May",
+        jun: "Jun",
+        jul: "Jul",
+        aug: "Aug",
+        sep: "Sep",
+        oct: "Oct",
+        nov: "Nov",
+        dec: "Dec"
+      }
+    },
+    inputNumber: {
+      decrease: "decrease number",
+      increase: "increase number"
+    },
+    select: {
+      loading: "Loading",
+      noMatch: "No matching data",
+      noData: "No data",
+      placeholder: "Select"
+    },
+    dropdown: {
+      toggleDropdown: "Toggle Dropdown"
+    },
+    cascader: {
+      noMatch: "No matching data",
+      loading: "Loading",
+      placeholder: "Select",
+      noData: "No data"
+    },
+    pagination: {
+      goto: "Go to",
+      pagesize: "/page",
+      total: "Total {total}",
+      pageClassifier: "",
+      page: "Page",
+      prev: "Go to previous page",
+      next: "Go to next page",
+      currentPage: "page {pager}",
+      prevPages: "Previous {pager} pages",
+      nextPages: "Next {pager} pages",
+      deprecationWarning: "Deprecated usages detected, please refer to the el-pagination documentation for more details"
+    },
+    dialog: {
+      close: "Close this dialog"
+    },
+    drawer: {
+      close: "Close this dialog"
+    },
+    messagebox: {
+      title: "Message",
+      confirm: "OK",
+      cancel: "Cancel",
+      error: "Illegal input",
+      close: "Close this dialog"
+    },
+    upload: {
+      deleteTip: "press delete to remove",
+      delete: "Delete",
+      preview: "Preview",
+      continue: "Continue"
+    },
+    slider: {
+      defaultLabel: "slider between {min} and {max}",
+      defaultRangeStartLabel: "pick start value",
+      defaultRangeEndLabel: "pick end value"
+    },
+    table: {
+      emptyText: "No Data",
+      confirmFilter: "Confirm",
+      resetFilter: "Reset",
+      clearFilter: "All",
+      sumText: "Sum"
+    },
+    tree: {
+      emptyText: "No Data"
+    },
+    transfer: {
+      noMatch: "No matching data",
+      noData: "No data",
+      titles: ["List 1", "List 2"],
+      filterPlaceholder: "Enter keyword",
+      noCheckedFormat: "{total} items",
+      hasCheckedFormat: "{checked}/{total} checked"
+    },
+    image: {
+      error: "FAILED"
+    },
+    pageHeader: {
+      title: "Back"
+    },
+    popconfirm: {
+      confirmButtonText: "Yes",
+      cancelButtonText: "No"
+    }
+  }
+};

apps/web/vue3/static/cdn/en.mjs

@@ -0,0 +1,165 @@
+/*! Element Plus v2.3.14 */
+
+var en = {
+  name: "en",
+  el: {
+    colorpicker: {
+      confirm: "OK",
+      clear: "Clear",
+      defaultLabel: "color picker",
+      description: "current color is {color}. press enter to select a new color."
+    },
+    datepicker: {
+      now: "Now",
+      today: "Today",
+      cancel: "Cancel",
+      clear: "Clear",
+      confirm: "OK",
+      dateTablePrompt: "Use the arrow keys and enter to select the day of the month",
+      monthTablePrompt: "Use the arrow keys and enter to select the month",
+      yearTablePrompt: "Use the arrow keys and enter to select the year",
+      selectedDate: "Selected date",
+      selectDate: "Select date",
+      selectTime: "Select time",
+      startDate: "Start Date",
+      startTime: "Start Time",
+      endDate: "End Date",
+      endTime: "End Time",
+      prevYear: "Previous Year",
+      nextYear: "Next Year",
+      prevMonth: "Previous Month",
+      nextMonth: "Next Month",
+      year: "",
+      month1: "January",
+      month2: "February",
+      month3: "March",
+      month4: "April",
+      month5: "May",
+      month6: "June",
+      month7: "July",
+      month8: "August",
+      month9: "September",
+      month10: "October",
+      month11: "November",
+      month12: "December",
+      week: "week",
+      weeks: {
+        sun: "Sun",
+        mon: "Mon",
+        tue: "Tue",
+        wed: "Wed",
+        thu: "Thu",
+        fri: "Fri",
+        sat: "Sat"
+      },
+      weeksFull: {
+        sun: "Sunday",
+        mon: "Monday",
+        tue: "Tuesday",
+        wed: "Wednesday",
+        thu: "Thursday",
+        fri: "Friday",
+        sat: "Saturday"
+      },
+      months: {
+        jan: "Jan",
+        feb: "Feb",
+        mar: "Mar",
+        apr: "Apr",
+        may: "May",
+        jun: "Jun",
+        jul: "Jul",
+        aug: "Aug",
+        sep: "Sep",
+        oct: "Oct",
+        nov: "Nov",
+        dec: "Dec"
+      }
+    },
+    inputNumber: {
+      decrease: "decrease number",
+      increase: "increase number"
+    },
+    select: {
+      loading: "Loading",
+      noMatch: "No matching data",
+      noData: "No data",
+      placeholder: "Select"
+    },
+    dropdown: {
+      toggleDropdown: "Toggle Dropdown"
+    },
+    cascader: {
+      noMatch: "No matching data",
+      loading: "Loading",
+      placeholder: "Select",
+      noData: "No data"
+    },
+    pagination: {
+      goto: "Go to",
+      pagesize: "/page",
+      total: "Total {total}",
+      pageClassifier: "",
+      page: "Page",
+      prev: "Go to previous page",
+      next: "Go to next page",
+      currentPage: "page {pager}",
+      prevPages: "Previous {pager} pages",
+      nextPages: "Next {pager} pages",
+      deprecationWarning: "Deprecated usages detected, please refer to the el-pagination documentation for more details"
+    },
+    dialog: {
+      close: "Close this dialog"
+    },
+    drawer: {
+      close: "Close this dialog"
+    },
+    messagebox: {
+      title: "Message",
+      confirm: "OK",
+      cancel: "Cancel",
+      error: "Illegal input",
+      close: "Close this dialog"
+    },
+    upload: {
+      deleteTip: "press delete to remove",
+      delete: "Delete",
+      preview: "Preview",
+      continue: "Continue"
+    },
+    slider: {
+      defaultLabel: "slider between {min} and {max}",
+      defaultRangeStartLabel: "pick start value",
+      defaultRangeEndLabel: "pick end value"
+    },
+    table: {
+      emptyText: "No Data",
+      confirmFilter: "Confirm",
+      resetFilter: "Reset",
+      clearFilter: "All",
+      sumText: "Sum"
+    },
+    tree: {
+      emptyText: "No Data"
+    },
+    transfer: {
+      noMatch: "No matching data",
+      noData: "No data",
+      titles: ["List 1", "List 2"],
+      filterPlaceholder: "Enter keyword",
+      noCheckedFormat: "{total} items",
+      hasCheckedFormat: "{checked}/{total} checked"
+    },
+    image: {
+      error: "FAILED"
+    },
+    pageHeader: {
+      title: "Back"
+    },
+    popconfirm: {
+      confirmButtonText: "Yes",
+      cancelButtonText: "No"
+    }
+  }
+};
+

apps/web/vue3/static/cdn/zh-cn.js

@@ -0,0 +1,130 @@
+// global wrapper
+/*! Element Plus v2.3.14 */
+
+window.zhCn = {
+  name: "zh-cn",
+  el: {
+    colorpicker: {
+      confirm: "\u786E\u5B9A",
+      clear: "\u6E05\u7A7A"
+    },
+    datepicker: {
+      now: "\u6B64\u523B",
+      today: "\u4ECA\u5929",
+      cancel: "\u53D6\u6D88",
+      clear: "\u6E05\u7A7A",
+      confirm: "\u786E\u5B9A",
+      selectDate: "\u9009\u62E9\u65E5\u671F",
+      selectTime: "\u9009\u62E9\u65F6\u95F4",
+      startDate: "\u5F00\u59CB\u65E5\u671F",
+      startTime: "\u5F00\u59CB\u65F6\u95F4",
+      endDate: "\u7ED3\u675F\u65E5\u671F",
+      endTime: "\u7ED3\u675F\u65F6\u95F4",
+      prevYear: "\u524D\u4E00\u5E74",
+      nextYear: "\u540E\u4E00\u5E74",
+      prevMonth: "\u4E0A\u4E2A\u6708",
+      nextMonth: "\u4E0B\u4E2A\u6708",
+      year: "\u5E74",
+      month1: "1 \u6708",
+      month2: "2 \u6708",
+      month3: "3 \u6708",
+      month4: "4 \u6708",
+      month5: "5 \u6708",
+      month6: "6 \u6708",
+      month7: "7 \u6708",
+      month8: "8 \u6708",
+      month9: "9 \u6708",
+      month10: "10 \u6708",
+      month11: "11 \u6708",
+      month12: "12 \u6708",
+      weeks: {
+        sun: "\u65E5",
+        mon: "\u4E00",
+        tue: "\u4E8C",
+        wed: "\u4E09",
+        thu: "\u56DB",
+        fri: "\u4E94",
+        sat: "\u516D"
+      },
+      months: {
+        jan: "\u4E00\u6708",
+        feb: "\u4E8C\u6708",
+        mar: "\u4E09\u6708",
+        apr: "\u56DB\u6708",
+        may: "\u4E94\u6708",
+        jun: "\u516D\u6708",
+        jul: "\u4E03\u6708",
+        aug: "\u516B\u6708",
+        sep: "\u4E5D\u6708",
+        oct: "\u5341\u6708",
+        nov: "\u5341\u4E00\u6708",
+        dec: "\u5341\u4E8C\u6708"
+      }
+    },
+    select: {
+      loading: "\u52A0\u8F7D\u4E2D",
+      noMatch: "\u65E0\u5339\u914D\u6570\u636E",
+      noData: "\u65E0\u6570\u636E",
+      placeholder: "\u8BF7\u9009\u62E9"
+    },
+    cascader: {
+      noMatch: "\u65E0\u5339\u914D\u6570\u636E",
+      loading: "\u52A0\u8F7D\u4E2D",
+      placeholder: "\u8BF7\u9009\u62E9",
+      noData: "\u6682\u65E0\u6570\u636E"
+    },
+    pagination: {
+      goto: "\u524D\u5F80",
+      pagesize: "\u6761/\u9875",
+      total: "\u5171 {total} \u6761",
+      pageClassifier: "\u9875",
+      page: "\u9875",
+      prev: "\u4E0A\u4E00\u9875",
+      next: "\u4E0B\u4E00\u9875",
+      currentPage: "\u7B2C {pager} \u9875",
+      prevPages: "\u5411\u524D {pager} \u9875",
+      nextPages: "\u5411\u540E {pager} \u9875",
+      deprecationWarning: "\u4F60\u4F7F\u7528\u4E86\u4E00\u4E9B\u5DF2\u88AB\u5E9F\u5F03\u7684\u7528\u6CD5\uFF0C\u8BF7\u53C2\u8003 el-pagination \u7684\u5B98\u65B9\u6587\u6863"
+    },
+    messagebox: {
+      title: "\u63D0\u793A",
+      confirm: "\u786E\u5B9A",
+      cancel: "\u53D6\u6D88",
+      error: "\u8F93\u5165\u7684\u6570\u636E\u4E0D\u5408\u6CD5!"
+    },
+    upload: {
+      deleteTip: "\u6309 delete \u952E\u53EF\u5220\u9664",
+      delete: "\u5220\u9664",
+      preview: "\u67E5\u770B\u56FE\u7247",
+      continue: "\u7EE7\u7EED\u4E0A\u4F20"
+    },
+    table: {
+      emptyText: "\u6682\u65E0\u6570\u636E",
+      confirmFilter: "\u7B5B\u9009",
+      resetFilter: "\u91CD\u7F6E",
+      clearFilter: "\u5168\u90E8",
+      sumText: "\u5408\u8BA1"
+    },
+    tree: {
+      emptyText: "\u6682\u65E0\u6570\u636E"
+    },
+    transfer: {
+      noMatch: "\u65E0\u5339\u914D\u6570\u636E",
+      noData: "\u65E0\u6570\u636E",
+      titles: ["\u5217\u8868 1", "\u5217\u8868 2"],
+      filterPlaceholder: "\u8BF7\u8F93\u5165\u641C\u7D22\u5185\u5BB9",
+      noCheckedFormat: "\u5171 {total} \u9879",
+      hasCheckedFormat: "\u5DF2\u9009 {checked}/{total} \u9879"
+    },
+    image: {
+      error: "\u52A0\u8F7D\u5931\u8D25"
+    },
+    pageHeader: {
+      title: "\u8FD4\u56DE"
+    },
+    popconfirm: {
+      confirmButtonText: "\u786E\u5B9A",
+      cancelButtonText: "\u53D6\u6D88"
+    }
+  }
+};

apps/web/vue3/static/cdn/zh-cn.mjs

@@ -0,0 +1,130 @@
+/*! Element Plus v2.3.14 */
+
+var zhCn = {
+  name: "zh-cn",
+  el: {
+    colorpicker: {
+      confirm: "\u786E\u5B9A",
+      clear: "\u6E05\u7A7A"
+    },
+    datepicker: {
+      now: "\u6B64\u523B",
+      today: "\u4ECA\u5929",
+      cancel: "\u53D6\u6D88",
+      clear: "\u6E05\u7A7A",
+      confirm: "\u786E\u5B9A",
+      selectDate: "\u9009\u62E9\u65E5\u671F",
+      selectTime: "\u9009\u62E9\u65F6\u95F4",
+      startDate: "\u5F00\u59CB\u65E5\u671F",
+      startTime: "\u5F00\u59CB\u65F6\u95F4",
+      endDate: "\u7ED3\u675F\u65E5\u671F",
+      endTime: "\u7ED3\u675F\u65F6\u95F4",
+      prevYear: "\u524D\u4E00\u5E74",
+      nextYear: "\u540E\u4E00\u5E74",
+      prevMonth: "\u4E0A\u4E2A\u6708",
+      nextMonth: "\u4E0B\u4E2A\u6708",
+      year: "\u5E74",
+      month1: "1 \u6708",
+      month2: "2 \u6708",
+      month3: "3 \u6708",
+      month4: "4 \u6708",
+      month5: "5 \u6708",
+      month6: "6 \u6708",
+      month7: "7 \u6708",
+      month8: "8 \u6708",
+      month9: "9 \u6708",
+      month10: "10 \u6708",
+      month11: "11 \u6708",
+      month12: "12 \u6708",
+      weeks: {
+        sun: "\u65E5",
+        mon: "\u4E00",
+        tue: "\u4E8C",
+        wed: "\u4E09",
+        thu: "\u56DB",
+        fri: "\u4E94",
+        sat: "\u516D"
+      },
+      months: {
+        jan: "\u4E00\u6708",
+        feb: "\u4E8C\u6708",
+        mar: "\u4E09\u6708",
+        apr: "\u56DB\u6708",
+        may: "\u4E94\u6708",
+        jun: "\u516D\u6708",
+        jul: "\u4E03\u6708",
+        aug: "\u516B\u6708",
+        sep: "\u4E5D\u6708",
+        oct: "\u5341\u6708",
+        nov: "\u5341\u4E00\u6708",
+        dec: "\u5341\u4E8C\u6708"
+      }
+    },
+    select: {
+      loading: "\u52A0\u8F7D\u4E2D",
+      noMatch: "\u65E0\u5339\u914D\u6570\u636E",
+      noData: "\u65E0\u6570\u636E",
+      placeholder: "\u8BF7\u9009\u62E9"
+    },
+    cascader: {
+      noMatch: "\u65E0\u5339\u914D\u6570\u636E",
+      loading: "\u52A0\u8F7D\u4E2D",
+      placeholder: "\u8BF7\u9009\u62E9",
+      noData: "\u6682\u65E0\u6570\u636E"
+    },
+    pagination: {
+      goto: "\u524D\u5F80",
+      pagesize: "\u6761/\u9875",
+      total: "\u5171 {total} \u6761",
+      pageClassifier: "\u9875",
+      page: "\u9875",
+      prev: "\u4E0A\u4E00\u9875",
+      next: "\u4E0B\u4E00\u9875",
+      currentPage: "\u7B2C {pager} \u9875",
+      prevPages: "\u5411\u524D {pager} \u9875",
+      nextPages: "\u5411\u540E {pager} \u9875",
+      deprecationWarning: "\u4F60\u4F7F\u7528\u4E86\u4E00\u4E9B\u5DF2\u88AB\u5E9F\u5F03\u7684\u7528\u6CD5\uFF0C\u8BF7\u53C2\u8003 el-pagination \u7684\u5B98\u65B9\u6587\u6863"
+    },
+    messagebox: {
+      title: "\u63D0\u793A",
+      confirm: "\u786E\u5B9A",
+      cancel: "\u53D6\u6D88",
+      error: "\u8F93\u5165\u7684\u6570\u636E\u4E0D\u5408\u6CD5!"
+    },
+    upload: {
+      deleteTip: "\u6309 delete \u952E\u53EF\u5220\u9664",
+      delete: "\u5220\u9664",
+      preview: "\u67E5\u770B\u56FE\u7247",
+      continue: "\u7EE7\u7EED\u4E0A\u4F20"
+    },
+    table: {
+      emptyText: "\u6682\u65E0\u6570\u636E",
+      confirmFilter: "\u7B5B\u9009",
+      resetFilter: "\u91CD\u7F6E",
+      clearFilter: "\u5168\u90E8",
+      sumText: "\u5408\u8BA1"
+    },
+    tree: {
+      emptyText: "\u6682\u65E0\u6570\u636E"
+    },
+    transfer: {
+      noMatch: "\u65E0\u5339\u914D\u6570\u636E",
+      noData: "\u65E0\u6570\u636E",
+      titles: ["\u5217\u8868 1", "\u5217\u8868 2"],
+      filterPlaceholder: "\u8BF7\u8F93\u5165\u641C\u7D22\u5185\u5BB9",
+      noCheckedFormat: "\u5171 {total} \u9879",
+      hasCheckedFormat: "\u5DF2\u9009 {checked}/{total} \u9879"
+    },
+    image: {
+      error: "\u52A0\u8F7D\u5931\u8D25"
+    },
+    pageHeader: {
+      title: "\u8FD4\u56DE"
+    },
+    popconfirm: {
+      confirmButtonText: "\u786E\u5B9A",
+      cancelButtonText: "\u53D6\u6D88"
+    }
+  }
+};
+

apps/web/vue3/static/conversation-sidebar.js

@@ -1,373 +1,373 @@
 /**
  * Conversation Sidebar Component
- * 类似ChatGPT的会话侧边栏组件
+ * Conversation Sidebar component
  */
 
 const ConversationSidebar = {
-    template: `
-        <div class="conversation-sidebar" :class="{ collapsed: isCollapsed }">
-            <!-- 头部：New Chat按钮 -->
-            <div class="sidebar-header">
-                <button class="new-chat-btn" @click="createNewConversation">
-                    <el-icon><plus /></el-icon>
-                    <span v-if="!isCollapsed">New Chat</span>
-                </button>
-                <button class="toggle-btn" @click="toggleSidebar">
-                    <el-icon><arrow-left v-if="!isCollapsed" /><arrow-right v-else /></el-icon>
-                </button>
-            </div>
-            
-            <!-- 搜索框 -->
-            <div class="search-box" v-if="!isCollapsed">
-                <el-input
-                    v-model="searchQuery"
-                    placeholder="搜索会话..."
-                    prefix-icon="search"
-                    clearable
-                    size="small"
-                />
-            </div>
-            
-            <!-- 会话列表 -->
-            <div class="conversation-list" v-if="!isCollapsed">
-                <!-- 今日会话 -->
-                <div class="list-section" v-if="todayConversations.length > 0">
-                    <div class="section-title">今天</div>
-                    <div
-                        v-for="conv in todayConversations"
-                        :key="conv.conversation_id"
-                        class="conversation-item"
-                        :class="{ active: currentConversationId === conv.conversation_id }"
-                        @click="selectConversation(conv)"
-                    >
-                        <div class="item-icon">
-                            <el-icon><chat-dot-round /></el-icon>
-                        </div>
-                        <div class="item-content">
-                            <div class="item-title">{{ conv.title }}</div>
-                            <div class="item-meta">
-                                <span class="gene-family" v-if="conv.gene_family">{{ conv.gene_family }}</span>
-                                <span class="message-count">{{ conv.message_count }}条消息</span>
-                            </div>
-                            <div class="item-preview">{{ conv.last_message }}</div>
-                        </div>
-                        <div class="item-actions" @click.stop>
-                            <el-dropdown trigger="click">
-                                <el-icon><more /></el-icon>
-                                <template #dropdown>
-                                    <el-dropdown-menu>
-                                        <el-dropdown-item @click="renameConversation(conv)">
-                                            <el-icon><edit /></el-icon>重命名
-                                        </el-dropdown-item>
-                                        <el-dropdown-item @click="archiveConversation(conv)">
-                                            <el-icon><folder /></el-icon>归档
-                                        </el-dropdown-item>
-                                        <el-dropdown-item divided @click="deleteConversation(conv)" type="danger">
-                                            <el-icon><delete /></el-icon>删除
-                                        </el-dropdown-item>
-                                    </el-dropdown-menu>
-                                </template>
-                            </el-dropdown>
-                        </div>
-                    </div>
-                </div>
-                
-                <!-- 昨日会话 -->
-                <div class="list-section" v-if="yesterdayConversations.length > 0">
-                    <div class="section-title">昨天</div>
-                    <div
-                        v-for="conv in yesterdayConversations"
-                        :key="conv.conversation_id"
-                        class="conversation-item"
-                        :class="{ active: currentConversationId === conv.conversation_id }"
-                        @click="selectConversation(conv)"
-                    >
-                        <div class="item-icon">
-                            <el-icon><chat-dot-round /></el-icon>
-                        </div>
-                        <div class="item-content">
-                            <div class="item-title">{{ conv.title }}</div>
-                            <div class="item-meta">
-                                <span class="gene-family" v-if="conv.gene_family">{{ conv.gene_family }}</span>
-                                <span class="message-count">{{ conv.message_count }}条</span>
-                            </div>
-                        </div>
-                    </div>
-                </div>
-                
-                <!-- 更早会话 -->
-                <div class="list-section" v-if="olderConversations.length > 0">
-                    <div class="section-title">过去7天</div>
-                    <div
-                        v-for="conv in olderConversations"
-                        :key="conv.conversation_id"
-                        class="conversation-item"
-                        :class="{ active: currentConversationId === conv.conversation_id }"
-                        @click="selectConversation(conv)"
-                    >
-                        <div class="item-icon">
-                            <el-icon><chat-dot-round /></el-icon>
-                        </div>
-                        <div class="item-content">
-                            <div class="item-title">{{ conv.title }}</div>
-                            <div class="item-meta">
-                                <span class="gene-family" v-if="conv.gene_family">{{ conv.gene_family }}</span>
-                            </div>
-                        </div>
-                    </div>
-                </div>
-                
-                <!-- 空状态 -->
-                <div v-if="filteredConversations.length === 0" class="empty-state">
-                    <el-icon><chat-line-round /></el-icon>
-                    <p>暂无会话</p>
-                    <p class="hint">点击上方"New Chat"开始新的基因家族分析</p>
-                </div>
-            </div>
-            
-            <!-- 底部：用户信息 -->
-            <div class="sidebar-footer" v-if="!isCollapsed">
-                <div class="user-info">
-                    <el-avatar :size="32" :icon="UserFilled" />
-                    <span class="username">{{ userId }}</span>
-                </div>
-            </div>
-        </div>
-    `,
-    
-    props: {
-        userId: {
-            type: String,
-            default: 'xia-lab'
-        },
-        apiBaseUrl: {
-            type: String,
-            default: 'http://localhost:8000/api'
-        }
-    },
-    
-    data() {
-        return {
-            isCollapsed: false,
-            searchQuery: '',
-            conversations: [],
-            currentConversationId: null,
-            loading: false,
-            // 图标
-            Plus: ElementPlusIconsVue.Plus,
-            ArrowLeft: ElementPlusIconsVue.ArrowLeft,
-            ArrowRight: ElementPlusIconsVue.ArrowRight,
-            ChatDotRound: ElementPlusIconsVue.ChatDotRound,
-            ChatLineRound: ElementPlusIconsVue.ChatLineRound,
-            More: ElementPlusIconsVue.More,
-            Edit: ElementPlusIconsVue.Edit,
-            Folder: ElementPlusIconsVue.Folder,
-            Delete: ElementPlusIconsVue.Delete,
-            UserFilled: ElementPlusIconsVue.UserFilled,
-        }
-    },
-    
-    computed: {
-        filteredConversations() {
-            if (!this.searchQuery) return this.conversations;
-            const query = this.searchQuery.toLowerCase();
-            return this.conversations.filter(conv => 
-                conv.title.toLowerCase().includes(query) ||
-                (conv.gene_family && conv.gene_family.toLowerCase().includes(query)) ||
-                conv.last_message.toLowerCase().includes(query)
-            );
-        },
-        
-        todayConversations() {
-            const today = new Date().toDateString();
-            return this.filteredConversations.filter(conv => {
-                const convDate = new Date(conv.updated_at).toDateString();
-                return convDate === today;
-            });
-        },
-        
-        yesterdayConversations() {
-            const yesterday = new Date(Date.now() - 86400000).toDateString();
-            return this.filteredConversations.filter(conv => {
-                const convDate = new Date(conv.updated_at).toDateString();
-                return convDate === yesterday;
-            });
-        },
-        
-        olderConversations() {
-            const yesterday = new Date(Date.now() - 86400000).toDateString();
-            return this.filteredConversations.filter(conv => {
-                const convDate = new Date(conv.updated_at).toDateString();
-                return convDate < yesterday;
-            });
-        }
-    },
-    
-    mounted() {
-        this.loadConversations();
-        // 定时刷新
-        this.refreshInterval = setInterval(() => {
-            this.loadConversations();
-        }, 30000); // 30秒刷新一次
-    },
-    
-    beforeUnmount() {
-        if (this.refreshInterval) {
-            clearInterval(this.refreshInterval);
-        }
-    },
-    
-    methods: {
-        async loadConversations() {
-            try {
-                this.loading = true;
-                const response = await fetch(`${this.apiBaseUrl}/conversations/list/${this.userId}`);
-                if (response.ok) {
-                    const data = await response.json();
-                    this.conversations = data.conversations || [];
-                }
-            } catch (error) {
-                console.error('Failed to load conversations:', error);
-                // 使用模拟数据
-                this.conversations = this.getMockConversations();
-            } finally {
-                this.loading = false;
-            }
-        },
-        
-        getMockConversations() {
-            // 模拟数据，用于测试
-            return [
-                {
-                    conversation_id: 'conv_001',
-                    title: 'At.WRKY基因家族分析',
-                    gene_family: 'WRKY',
-                    species: 'At',
-                    message_count: 12,
-                    last_message: '系统发育树已生成完成',
-                    updated_at: new Date().toISOString()
-                },
-                {
-                    conversation_id: 'conv_002',
-                    title: 'Os.TCP进化分析',
-                    gene_family: 'TCP',
-                    species: 'Os',
-                    message_count: 5,
-                    last_message: '正在构建进化树...',
-                    updated_at: new Date(Date.now() - 3600000).toISOString()
-                },
-                {
-                    conversation_id: 'conv_003',
-                    title: 'HMMER参数配置咨询',
-                    message_count: 8,
-                    last_message: 'E-value建议设置为1e-5',
-                    updated_at: new Date(Date.now() - 86400000).toISOString()
-                }
-            ];
-        },
-        
-        toggleSidebar() {
-            this.isCollapsed = !this.isCollapsed;
-            this.$emit('toggle', this.isCollapsed);
-        },
-        
-        async createNewConversation() {
-            try {
-                const response = await fetch(`${this.apiBaseUrl}/conversations/create`, {
-                    method: 'POST',
-                    headers: { 'Content-Type': 'application/json' },
-                    body: JSON.stringify({
-                        user_id: this.userId,
-                        title: '新会话 ' + new Date().toLocaleTimeString()
-                    })
-                });
-                
-                if (response.ok) {
-                    const data = await response.json();
-                    this.currentConversationId = data.conversation_id;
-                    this.$emit('new-conversation', data);
-                    await this.loadConversations();
-                }
-            } catch (error) {
-                console.error('Failed to create conversation:', error);
-                // 创建本地会话
-                const newConv = {
-                    conversation_id: 'conv_' + Date.now(),
-                    title: '新会话',
-                    message_count: 0,
-                    last_message: '',
-                    updated_at: new Date().toISOString()
-                };
-                this.conversations.unshift(newConv);
-                this.currentConversationId = newConv.conversation_id;
-                this.$emit('new-conversation', newConv);
-            }
-        },
-        
-        selectConversation(conv) {
-            this.currentConversationId = conv.conversation_id;
-            this.$emit('select-conversation', conv);
-        },
-        
-        async renameConversation(conv) {
-            const newTitle = prompt('请输入新标题:', conv.title);
-            if (newTitle && newTitle !== conv.title) {
-                try {
-                    await fetch(`${this.apiBaseUrl}/conversations/update-title`, {
-                        method: 'POST',
-                        headers: { 'Content-Type': 'application/json' },
-                        body: JSON.stringify({
-                            conversation_id: conv.conversation_id,
-                            title: newTitle
-                        })
-                    });
-                    conv.title = newTitle;
-                } catch (error) {
-                    console.error('Failed to rename:', error);
-                    conv.title = newTitle; // 本地更新
-                }
-            }
-        },
-        
-        async archiveConversation(conv) {
-            try {
-                await fetch(`${this.apiBaseUrl}/conversations/archive`, {
-                    method: 'POST',
-                    headers: { 'Content-Type': 'application/json' },
-                    body: JSON.stringify({
-                        conversation_id: conv.conversation_id
-                    })
-                });
-                await this.loadConversations();
-            } catch (error) {
-                console.error('Failed to archive:', error);
-            }
-        },
-        
-        async deleteConversation(conv) {
-            if (confirm(`确定要删除会话 "${conv.title}" 吗？`)) {
-                try {
-                    await fetch(`${this.apiBaseUrl}/conversations/delete`, {
-                        method: 'POST',
-                        headers: { 'Content-Type': 'application/json' },
-                        body: JSON.stringify({
-                            conversation_id: conv.conversation_id
-                        })
-                    });
-                    await this.loadConversations();
-                    if (this.currentConversationId === conv.conversation_id) {
-                        this.currentConversationId = null;
-                        this.$emit('select-conversation', null);
-                    }
-                } catch (error) {
-                    console.error('Failed to delete:', error);
-                }
-            }
-        }
-    }
+ template: `
+ <div class="conversation-sidebar" :class="{ collapsed: isCollapsed }">
+ <!-- Header: New Chat button -->
+ <div class="sidebar-header">
+ <button class="new-chat-btn" @click="createNewConversation">
+ <el-icon><plus /></el-icon>
+ <span v-if="!isCollapsed">New Chat</span>
+ </button>
+ <button class="toggle-btn" @click="toggleSidebar">
+ <el-icon><arrow-left v-if="!isCollapsed" /><arrow-right v-else /></el-icon>
+ </button>
+ </div>
+ 
+ <!-- Search box -->
+ <div class="search-box" v-if="!isCollapsed">
+ <el-input
+ v-model="searchQuery"
+ placeholder="Search conversations..."
+ prefix-icon="search"
+ clearable
+ size="small"
+ />
+ </div>
+ 
+ <!-- Conversation list -->
+ <div class="conversation-list" v-if="!isCollapsed">
+ <!-- Todays conversations -->
+ <div class="list-section" v-if="todayConversations.length > 0">
+ <div class="section-title">Today</div>
+ <div
+ v-for="conv in todayConversations"
+ :key="conv.conversation_id"
+ class="conversation-item"
+ :class="{ active: currentConversationId === conv.conversation_id }"
+ @click="selectConversation(conv)"
+ >
+ <div class="item-icon">
+ <el-icon><chat-dot-round /></el-icon>
+ </div>
+ <div class="item-content">
+ <div class="item-title">{{ conv.title }}</div>
+ <div class="item-meta">
+ <span class="gene-family" v-if="conv.gene_family">{{ conv.gene_family }}</span>
+ <span class="message-count">{{ conv.message_count }} messages</span>
+ </div>
+ <div class="item-preview">{{ conv.last_message }}</div>
+ </div>
+ <div class="item-actions" @click.stop>
+ <el-dropdown trigger="click">
+ <el-icon><more /></el-icon>
+ <template #dropdown>
+ <el-dropdown-menu>
+ <el-dropdown-item @click="renameConversation(conv)">
+ <el-icon><edit /></el-icon>Rename
+ </el-dropdown-item>
+ <el-dropdown-item @click="archiveConversation(conv)">
+ <el-icon><folder /></el-icon>Archive
+ </el-dropdown-item>
+ <el-dropdown-item divided @click="deleteConversation(conv)" type="danger">
+ <el-icon><delete /></el-icon>Delete
+ </el-dropdown-item>
+ </el-dropdown-menu>
+ </template>
+ </el-dropdown>
+ </div>
+ </div>
+ </div>
+ 
+ <!-- Yesterdays conversations -->
+ <div class="list-section" v-if="yesterdayConversations.length > 0">
+ <div class="section-title">Yesterday</div>
+ <div
+ v-for="conv in yesterdayConversations"
+ :key="conv.conversation_id"
+ class="conversation-item"
+ :class="{ active: currentConversationId === conv.conversation_id }"
+ @click="selectConversation(conv)"
+ >
+ <div class="item-icon">
+ <el-icon><chat-dot-round /></el-icon>
+ </div>
+ <div class="item-content">
+ <div class="item-title">{{ conv.title }}</div>
+ <div class="item-meta">
+ <span class="gene-family" v-if="conv.gene_family">{{ conv.gene_family }}</span>
+ <span class="message-count">{{ conv.message_count }}</span>
+ </div>
+ </div>
+ </div>
+ </div>
+ 
+ <!-- Older conversations -->
+ <div class="list-section" v-if="olderConversations.length > 0">
+ <div class="section-title">Past 7 days</div>
+ <div
+ v-for="conv in olderConversations"
+ :key="conv.conversation_id"
+ class="conversation-item"
+ :class="{ active: currentConversationId === conv.conversation_id }"
+ @click="selectConversation(conv)"
+ >
+ <div class="item-icon">
+ <el-icon><chat-dot-round /></el-icon>
+ </div>
+ <div class="item-content">
+ <div class="item-title">{{ conv.title }}</div>
+ <div class="item-meta">
+ <span class="gene-family" v-if="conv.gene_family">{{ conv.gene_family }}</span>
+ </div>
+ </div>
+ </div>
+ </div>
+ 
+ <!-- Empty state -->
+ <div v-if="filteredConversations.length === 0" class="empty-state">
+ <el-icon><chat-line-round /></el-icon>
+ <p>No conversations yet</p>
+ <p class="hint">Click "New Chat" to start gene family analysis</p>
+ </div>
+ </div>
+ 
+ <!-- footer: user info -->
+ <div class="sidebar-footer" v-if="!isCollapsed">
+ <div class="user-info">
+ <el-avatar :size="32" :icon="UserFilled" />
+ <span class="username">{{ userId }}</span>
+ </div>
+ </div>
+ </div>
+ `,
+ 
+ props: {
+ userId: {
+ type: String,
+ default: 'xia-lab'
+ },
+ apiBaseUrl: {
+ type: String,
+ default: 'http://localhost:8000/api'
+ }
+ },
+ 
+ data() {
+ return {
+ isCollapsed: false,
+ searchQuery: '',
+ conversations: [],
+ currentConversationId: null,
+ loading: false,
+ // Icons
+ Plus: ElementPlusIconsVue.Plus,
+ ArrowLeft: ElementPlusIconsVue.ArrowLeft,
+ ArrowRight: ElementPlusIconsVue.ArrowRight,
+ ChatDotRound: ElementPlusIconsVue.ChatDotRound,
+ ChatLineRound: ElementPlusIconsVue.ChatLineRound,
+ More: ElementPlusIconsVue.More,
+ Edit: ElementPlusIconsVue.Edit,
+ Folder: ElementPlusIconsVue.Folder,
+ Delete: ElementPlusIconsVue.Delete,
+ UserFilled: ElementPlusIconsVue.UserFilled,
+ }
+ },
+ 
+ computed: {
+ filteredConversations() {
+ if (!this.searchQuery) return this.conversations;
+ const query = this.searchQuery.toLowerCase();
+ return this.conversations.filter(conv => 
+ conv.title.toLowerCase().includes(query) ||
+ (conv.gene_family && conv.gene_family.toLowerCase().includes(query)) ||
+ conv.last_message.toLowerCase().includes(query)
+ );
+ },
+ 
+ todayConversations() {
+ const today = new Date().toDateString();
+ return this.filteredConversations.filter(conv => {
+ const convDate = new Date(conv.updated_at).toDateString();
+ return convDate === today;
+ });
+ },
+ 
+ yesterdayConversations() {
+ const yesterday = new Date(Date.now() - 86400000).toDateString();
+ return this.filteredConversations.filter(conv => {
+ const convDate = new Date(conv.updated_at).toDateString();
+ return convDate === yesterday;
+ });
+ },
+ 
+ olderConversations() {
+ const yesterday = new Date(Date.now() - 86400000).toDateString();
+ return this.filteredConversations.filter(conv => {
+ const convDate = new Date(conv.updated_at).toDateString();
+ return convDate < yesterday;
+ });
+ }
+ },
+ 
+ mounted() {
+ this.loadConversations();
+ // Auto-refresh
+ this.refreshInterval = setInterval(() => {
+ this.loadConversations();
+ }, 30000); // 30 seconds refresh interval
+ },
+ 
+ beforeUnmount() {
+ if (this.refreshInterval) {
+ clearInterval(this.refreshInterval);
+ }
+ },
+ 
+ methods: {
+ async loadConversations() {
+ try {
+ this.loading = true;
+ const response = await fetch(`${this.apiBaseUrl}/conversations/list/${this.userId}`);
+ if (response.ok) {
+ const data = await response.json();
+ this.conversations = data.conversations || [];
+ }
+ } catch (error) {
+ console.error('Failed to load conversations:', error);
+ // Use mock data
+ this.conversations = this.getMockConversations();
+ } finally {
+ this.loading = false;
+ }
+ },
+ 
+ getMockConversations() {
+ // Mock data for testing
+ return [
+ {
+ conversation_id: 'conv_001',
+ title: 'At.WRKY gene family analysis',
+ gene_family: 'WRKY',
+ species: 'At',
+ message_count: 12,
+ last_message: 'Phylogenetic tree done',
+ updated_at: new Date().toISOString()
+ },
+ {
+ conversation_id: 'conv_002',
+ title: 'Os.TCP evolution analysis',
+ gene_family: 'TCP',
+ species: 'Os',
+ message_count: 5,
+ last_message: 'Building tree...',
+ updated_at: new Date(Date.now() - 3600000).toISOString()
+ },
+ {
+ conversation_id: 'conv_003',
+ title: 'HMMER config consultation',
+ message_count: 8,
+ last_message: 'E-value 1e-5 recommended',
+ updated_at: new Date(Date.now() - 86400000).toISOString()
+ }
+ ];
+ },
+ 
+ toggleSidebar() {
+ this.isCollapsed = !this.isCollapsed;
+ this.$emit('toggle', this.isCollapsed);
+ },
+ 
+ async createNewConversation() {
+ try {
+ const response = await fetch(`${this.apiBaseUrl}/conversations/create`, {
+ method: 'POST',
+ headers: { 'Content-Type': 'application/json' },
+ body: JSON.stringify({
+ user_id: this.userId,
+ title: 'New Chat ' + new Date().toLocaleTimeString()
+ })
+ });
+ 
+ if (response.ok) {
+ const data = await response.json();
+ this.currentConversationId = data.conversation_id;
+ this.$emit('new-conversation', data);
+ await this.loadConversations();
+ }
+ } catch (error) {
+ console.error('Failed to create conversation:', error);
+ // Create local conversation
+ const newConv = {
+ conversation_id: 'conv_' + Date.now(),
+ title: 'New Chat',
+ message_count: 0,
+ last_message: '',
+ updated_at: new Date().toISOString()
+ };
+ this.conversations.unshift(newConv);
+ this.currentConversationId = newConv.conversation_id;
+ this.$emit('new-conversation', newConv);
+ }
+ },
+ 
+ selectConversation(conv) {
+ this.currentConversationId = conv.conversation_id;
+ this.$emit('select-conversation', conv);
+ },
+ 
+ async renameConversation(conv) {
+ const newTitle = prompt('Enter new title:', conv.title);
+ if (newTitle && newTitle !== conv.title) {
+ try {
+ await fetch(`${this.apiBaseUrl}/conversations/update-title`, {
+ method: 'POST',
+ headers: { 'Content-Type': 'application/json' },
+ body: JSON.stringify({
+ conversation_id: conv.conversation_id,
+ title: newTitle
+ })
+ });
+ conv.title = newTitle;
+ } catch (error) {
+ console.error('Failed to rename:', error);
+ conv.title = newTitle; // local update
+ }
+ }
+ },
+ 
+ async archiveConversation(conv) {
+ try {
+ await fetch(`${this.apiBaseUrl}/conversations/archive`, {
+ method: 'POST',
+ headers: { 'Content-Type': 'application/json' },
+ body: JSON.stringify({
+ conversation_id: conv.conversation_id
+ })
+ });
+ await this.loadConversations();
+ } catch (error) {
+ console.error('Failed to archive:', error);
+ }
+ },
+ 
+ async deleteConversation(conv) {
+ if (confirm(`Delete conversation "${conv.title}"?`)) {
+ try {
+ await fetch(`${this.apiBaseUrl}/conversations/delete`, {
+ method: 'POST',
+ headers: { 'Content-Type': 'application/json' },
+ body: JSON.stringify({
+ conversation_id: conv.conversation_id
+ })
+ });
+ await this.loadConversations();
+ if (this.currentConversationId === conv.conversation_id) {
+ this.currentConversationId = null;
+ this.$emit('select-conversation', null);
+ }
+ } catch (error) {
+ console.error('Failed to delete:', error);
+ }
+ }
+ }
+ }
 };
 
-// 导出组件
+// Export component
 if (typeof module !== 'undefined' && module.exports) {
-    module.exports = { ConversationSidebar };
+ module.exports = { ConversationSidebar };
 }

apps/web/vue3/static/i18n_en.js

@@ -0,0 +1,891 @@
+// ===== Agent 多语言：卡片 + Agent 消息英文翻译 =====
+// key 为中文原文，value 为英文翻译
+// 使用 t(key) 获取翻译，tp(key, params) 处理动态占位符
+
+window.__I18N_EN__ = {
+  "Agent状态": "AgentStatus",
+  "Self-RAG反思": "Self-RAG Reflection",
+  "三角形": "Triangle",
+  "下一步": "Next Step",
+  "下载": "Download",
+  "不受支持标记处理策略": "Unsupported Marker Handling Policy",
+  "中性替换率": "Neutral Substitution Rate",
+  "中性速率": "Neutral Rate",
+  "中文": "Chinese",
+  "丰度引擎": "Abundance Engine",
+  "主参考": "Primary Reference",
+  "主参考物种": "Primary reference species",
+  "主要结果": "Main Results",
+  "亚家族分类确认": "Subfamily Classification Confirmation",
+  "亚家族名称距离": "Subfamily Name Distance",
+  "亚家族外圈弧线": "Subfamily Outer Arc",
+  "亚家族确认参数": "Subfamily Confirmation Parameters",
+  "交集": "Intersection",
+  "仅圈图": "Circle Plot Only",
+  "仅比较物种": "Comparative Species Only",
+  "仅目标": "Target Only",
+  "仅目标物种": "Target Species Only",
+  "仅矩形图": "Rectangular Only",
+  "任务列表": "Task List",
+  "位置分布图": "Position Distribution Map",
+  "使用拟南芥作为显式兜底": "Use Arabidopsis as Explicit Fallback",
+  "使用推荐区间": "Use Recommended Range",
+  "保存方法学方案": "Save Methodology Plan",
+  "保留": "Retain",
+  "修订": "Revision",
+  "倒计时已暂停": "Countdown paused",
+  "全屏": "Full Screen",
+  "全屏展示": "Full Screen Display",
+  "全部": "All",
+  "全部行": "All Rows",
+  "关系总数": "Total Edges",
+  "关闭": "Close",
+  "内容": "Content",
+  "内置物种代码": "Built-in Species Code",
+  "分析家族": "Analysis Family",
+  "分组角色": "Group Role",
+  "分组角色不改": "Group Role (Unchanged)",
+  "删除": "Deleted",
+  "删除会话": "Delete Session",
+  "刷新": "Refresh",
+  "刷新结果": "Refresh Results",
+  "前景基因": "Foreground Genes",
+  "功能类别": "Feature Category",
+  "加入正式列表": "Add to Formal List",
+  "包含关键词": "Contains Keyword",
+  "匹配方式": "Match Mode",
+  "匹配行": "Matched Rows",
+  "原因": "Reason",
+  "原始代码": "Original Code",
+  "原始命令": "Original Command",
+  "原始文件名": "Original Filename",
+  "原始证据": "Original Evidence",
+  "原推荐参考": "Original Recommended Reference",
+  "参数": "Parameters",
+  "参考分类": "Reference Classification",
+  "参考前缀": "Reference Prefix",
+  "参考基因": "Reference Gene",
+  "参考基因编号": "Reference Gene ID",
+  "参考文献": "References",
+  "参考物种": "Reference Species",
+  "发送": "Send",
+  "取消": "Cancel",
+  "取消预览": "Cancel Preview",
+  "只会执行已勾选且可执行的阶段": "Only checked and executable stages will run",
+  "可改": "Editable",
+  "同时处理的样本数": "Number of Samples Processed Simultaneously",
+  "名称": "Name",
+  "后续主参考物种": "Downstream Primary Reference Species",
+  "启动子元件": "Promoter Elements",
+  "启动子自定义绘图": "Promoter Custom Plot",
+  "图例颜色": "Legend Color",
+  "图类型": "Plot Type",
+  "图表标题清单": "Chart Title List",
+  "图表标题驱动复现规划预览": "Chart Title-Driven Reproduction Plan Preview",
+  "圆形": "Circle",
+  "基因注释": "Gene Annotation",
+  "基本参数": "Basic Parameters",
+  "基础参数": "Basic Parameters",
+  "处理中...": "Processing...",
+  "处理方式": "Processing Method",
+  "多物种处理方式": "Multi-species Processing",
+  "大小": "Size",
+  "定量模式": "Quantification Mode",
+  "实验设计": "Experimental Design",
+  "实验设计模块": "Experimental Design Module",
+  "审核备注": "Review Notes",
+  "富集类型": "Enrichment Type",
+  "差异表达范围": "Differential Expression Scope",
+  "已写阶段": "Registered Stages",
+  "已完成": "Completed",
+  "已完成阶段": "Completed Stages",
+  "已应用": "Applied",
+  "建议参数": "Suggested Parameters",
+  "开始验证": "Start Validation",
+  "弧线避让距离": "Arc Avoidance Distance",
+  "当前参数": "Current Parameters",
+  "当前参考物种": "Current Reference Species",
+  "当前选择": "Current Selection",
+  "待补数据": "Awaiting Data",
+  "必填": "Required",
+  "我知道了": "Got It",
+  "手动区间": "Manual Range",
+  "执行": "Execute",
+  "扫描双链": "Scan Both Strands",
+  "批量执行确认": "Batch Execution Confirmation",
+  "拉丁名": "Latin Name",
+  "拓展分析运行参数确认": "Extension Analysis Run Parameter Confirmation",
+  "拓展分析阶段": "Extension Analysis Stage",
+  "拓展分析阶段状态": "Extension Analysis Stage Status",
+  "拓展分析阶段确认": "Extension Analysis Stage Confirmation",
+  "拓展分析阶段运行结果": "Extension Analysis Stage Run Result",
+  "按元件": "By Element",
+  "按单个物种分别分析与确认": "Per-Species Analysis & Confirmation",
+  "按当前筛选重建成员蛋白": "Rebuild Member Proteins by Current Filter",
+  "按类别": "By Category",
+  "授权发布": "Authorize Publish",
+  "推荐参考数据下载失败": "Recommended Ref Data Download Failed",
+  "提出修订意见": "Submit Revision Suggestions",
+  "操作": "Operation",
+  "数据准备": "Data Preparation",
+  "数据路径": "Data Path",
+  "数据需求": "Data Needs",
+  "文件": "File",
+  "文件类型": "File Type",
+  "文件范围": "File Scope",
+  "文献参数": "Literature Parameters",
+  "文献工具": "Literature Tools",
+  "文章涉及物种与下载入口": "Article Species & Download Entry",
+  "新增": "Added",
+  "方形": "Square",
+  "方法参考": "Method Reference",
+  "方法学文献": "Methodology Papers",
+  "无": "None",
+  "明确指定": "Explicitly Specify",
+  "显示": "Show",
+  "暂停倒计时": "Pause Countdown",
+  "暂无已上传文件": "No Uploaded Files Yet",
+  "暂无拓展分析阶段结果": "No Extension Analysis Stage Results Yet",
+  "暂无状态记录": "No Status Record Yet",
+  "最大宽度": "Max Width",
+  "最小宽度": "Min Width",
+  "最小非缺口比例": "Min Non-gap Ratio",
+  "最近到属": "Closest to Genus",
+  "最近到目": "Closest to Order",
+  "最近到科": "Closest to Family",
+  "最近到纲": "Closest to Class",
+  "有损转换警告": "Lossy Conversion Warning",
+  "服务器测试数据路径": "Server Test Data Path",
+  "未从图表标题推断出额外阶段": "No Additional Stages Inferred from Chart Titles",
+  "未提取到图表标题": "No Chart Title Extracted",
+  "未识别到内置下载链接": "No Built-in DL Link Identified",
+  "未识别到参数": "No Parameters Identified",
+  "未识别到文献工具": "No Literature Tool Identified",
+  "本次成员鉴定物种": "Current Member Identification Species",
+  "来源": "Sources",
+  "染色体标注位置": "Chromosome Annotation Position",
+  "染色体标注大小": "Chromosome Annotation Size",
+  "染色体策略": "Chromosome Strategy",
+  "染色体色块粗细": "Chromosome Block Width",
+  "查看本次知识检索过程": "View knowledge retrieval process",
+  "标记大小": "Mark Size",
+  "标记模式": "Mark Mode",
+  "标题推断的候选阶段": "Candidate Stages Inferred from Titles",
+  "树枝粗细": "Branch Width",
+  "树枝颜色": "Branch Color",
+  "样本并行数量": "Sample Parallel Count",
+  "检索结果": "Retrieval Results",
+  "检索表格内容": "Search Table Content",
+  "正则匹配": "Regex Match",
+  "此拓展阶段没有可配置参数": "No Configurable Parameters",
+  "测试数据": "Test Data",
+  "涉及阶段": "Involved Stages",
+  "添加行": "Add Row",
+  "灰色": "Gray",
+  "物种": "Species",
+  "物种代码": "Species Code",
+  "物种代码不改": "Species Code (Unchanged)",
+  "物种名": "Species Name",
+  "物种角色": "Species Role",
+  "物种角色不改": "Species Role (Unchanged)",
+  "环境": "Environment",
+  "留空则自动发现": "Leave empty for auto-discovery",
+  "目标前缀": "Target Prefix",
+  "目标物种": "Target Species",
+  "知识图谱": "Knowledge Graph",
+  "确定": "Confirm",
+  "确认亚家族标签": "Confirm Subfamily Labels",
+  "确认保存": "Confirm Save",
+  "确认并写入正式配置": "Confirm & Write to Official Config",
+  "确认并执行": "Confirm & Execute",
+  "确认执行": "Confirm Execution",
+  "确认自动注释": "Confirm Auto-Annotation",
+  "确认重绘": "Confirm Replot",
+  "种间共线性比较物种": "Interspecies Synteny Comparative Species",
+  "种间共线性绘图设置": "Interspecies Synteny Plot Settings",
+  "移除": "Remove",
+  "窗口长度": "Window Length",
+  "类型": "Types",
+  "精确匹配": "Exact Match",
+  "红色": "Red",
+  "终止位点": "End Site",
+  "终止残基": "End Residue",
+  "绘图模式": "Plot Mode",
+  "绿色": "Green",
+  "置信度": "Confidence",
+  "置信度阈值": "Confidence Threshold",
+  "群体遗传输入": "Population Genetic Input",
+  "背景来源": "Background Source",
+  "背景知识": "Background knowledge",
+  "自动": "Auto",
+  "自动从注释提取": "Auto-extract from Annotation",
+  "自动保守窗口": "Auto Conserved Window",
+  "自动判断": "Auto-detect",
+  "节点名称": "Node Name",
+  "节点总数": "Total Nodes",
+  "节点类型": "Node Types",
+  "菱形": "Diamond",
+  "蓝色": "Blue",
+  "蛋白": "Protein",
+  "蛋白文件": "Protein Files",
+  "表达矩阵": "Expression Matrix",
+  "诊断动作": "Diagnostic Actions",
+  "说明": "Description",
+  "请先试运行以记录真实输出": "Please Run a Trial First to Record Actual Output",
+  "请选择共线性分析类型": "Please select synteny analysis type",
+  "请选择模式物种": "Select Model Species",
+  "起始位点": "Start Site",
+  "起始残基": "Start Residue",
+  "跳数": "Hops",
+  "输出": "Output",
+  "输出前缀": "Output Prefix",
+  "输出标签": "Output Label",
+  "运行值": "Run Value",
+  "运行结果": "Run Result",
+  "进化树图修订参数": "Phylogenetic Tree Revision Parameters",
+  "连接关系数": "Connections",
+  "迭代": "Iteration",
+  "选择其他提取结果": "Select Other Extraction Result",
+  "选择工具": "Select Tool",
+  "选择已完成撰写的拓展阶段": "Select Completed Extension Stage",
+  "通常留空": "Usually leave empty",
+  "重写查询": "Rewrite Query",
+  "重新生成说明": "Regenerate Description",
+  "重置": "Reset",
+  "重置视图": "Reset View",
+  "链接": "Link",
+  "链特异性": "Library Strandedness",
+  "隐藏": "Hide",
+  "需检查": "Needs Review",
+  "需要补充": "Need to Supplement",
+  "非时间实验留空": "Leave empty for non-time experiments",
+  "项目": "Project",
+  "预览": "Preview",
+  "预览到表格": "Preview to Table",
+  "预览命令": "Preview Commands",
+  "黑色": "Black",
+  "默认值": "Default Value",
+  "提取方法学": "Extract",
+  "刷新记录": "Refresh",
+  "打开最新提取结果": "Open Latest Extraction",
+  "查看最新提取状态": "View Latest Status",
+  "打开所选提取结果": "Open Selected Extraction",
+  "查看所选提取状态": "View Selected Status",
+  "生成并发布": "Generate & Publish",
+  "运行已发布": "Run Published",
+  "当前会话": "Current Session",
+  "更多会话": "More Sessions",
+  "新建": "New",
+  "新建会话": "New Session",
+  "使用小Tip": "Tips",
+  "一、分析模块快速指令": "1. Analysis Module Quick Commands",
+  "选择输入框上方\"Analysis\"": "Select \"Analysis\" above input box",
+  "粘贴数据位置": "Paste data location",
+  "核对配置文件": "Review config files",
+  "执行数据准备": "Run data preparation",
+  "执行XX物种的亚家族分类": "Run subfamily classification for XX species",
+  "执行XX物种的基因家族分析": "Run gene family analysis for XX species",
+  "执行rnaseq分析": "Run RNA-seq analysis",
+  "二、文章复现步骤": "2. Article Reproduction Steps",
+  "点击\"浏览\"上传文献": "Click \"Browse\" to upload paper",
+  "点击提取方法学": "Click Extract Methodology",
+  "核对分析步骤": "Review analysis steps",
+  "点击保存方案": "Click Save Plan",
+  "上传数据": "Upload data",
+  "发送\"按已保存的方法学方案继续执行\"": "Send \"Continue with saved methodology plan\"",
+  "三、RAG快速指令": "3. RAG Quick Commands",
+  "选择输入框上方\"RAG\"": "Select \"RAG\" above input box",
+  "输入问题": "Enter question",
+  "点击\"知识图谱\"查看关系": "Click \"Knowledge Graph\" to view relations",
+  "知识图谱": "Knowledge Graph",
+  "重新加载": "Reload",
+  "重置视图": "Reset View",
+  "2.2 MSA Replot / 区间重绘": "2.2 MSA Replot",
+  "2.3 Tree Replot / 进化树重绘": "2.3 Tree Replot",
+  "2.5 亚家族确认": "2.5 Subfamily Confirmation",
+  "6.1 启动子自定义绘图": "6.1 Promoter Custom Plot",
+  "7️⃣ 种内Synteny": "7️⃣ Intraspecies Synteny",
+  "7.1 种内共线性 Circos 重绘": "7.1 Intraspecies Synteny Circos Replot",
+  "8️⃣ Karyotype": "8️⃣ Karyotype",
+  "9️⃣ 种间Synteny": "9️⃣ Interspecies Synteny",
+  "基因家族": "Gene Family",
+  "WGS2GWAS（遗传变异与群体分析）": "WGS2GWAS",
+  "基因家族 promoter 扫描": "Gene family promoter scanning",
+  "新会话": "New Session",
+  "新会话已创建，请选择或创建 Working Directory": "New session created. Please select or create a Working Directory",
+  "创建会话失败": "Failed to create session",
+  "会话已删除": "Session deleted",
+  "删除会话失败": "Failed to delete session",
+  "对话历史已清空": "Chat history cleared",
+  "错误日志已清空": "Error logs cleared",
+  "重启 Celery": "Restart Celery",
+  "重启项目": "Restart Project",
+  "提交配置好的项目重启脚本": "Submit configured project restart script",
+  "未配置 PLANTAI_RESTART_SCRIPT 或 restart_plantai_9000.sh": "PLANTAI_RESTART_SCRIPT or restart_plantai_9000.sh not configured",
+  "未配置项目重启脚本": "Project restart script not configured",
+  "这会提交项目重启脚本，当前页面可能短暂断开。": "This will submit the project restart script. The page may briefly disconnect.",
+  "确认重启": "Confirm Restart",
+  "项目重启脚本已提交": "Project restart script submitted",
+  "项目重启未提交": "Project restart not submitted",
+  "基因家族鉴定": "Gene Family Identification",
+  "系统发育分析": "Phylogenetic Analysis",
+  "Motif分析": "Motif Analysis",
+  "结构域分析": "Domain Analysis",
+  "基因结构分析": "Gene Structure Analysis",
+  "启动子元件预测": "Promoter Element Prediction",
+  "TF结合位点预测": "TF Binding Site Prediction",
+  "种内共线性分析": "Intraspecies Synteny Analysis",
+  "种间共线性分析": "Interspecies Synteny Analysis",
+  "核型/染色体分布分析": "Karyotype/Chromosome Distribution",
+  "PPI分析": "PPI Analysis",
+  "Ka/Ks分析": "Ka/Ks Analysis",
+  "蛋白理化性质与亚细胞定位": "Protein Properties & Subcellular Localization",
+  "表达矩阵分析": "Expression Matrix Analysis",
+  "RNA-seq QC": "RNA-seq QC",
+  "RNA-seq Align": "RNA-seq Align",
+  "RNA-seq Quant": "RNA-seq Quant",
+  "RNA-seq Sample QC": "RNA-seq Sample QC",
+  "GWAS VCF排序与HapMap转换": "GWAS VCF Sort & HapMap Conversion",
+  "GWAS TASSEL亲缘关系矩阵": "GWAS TASSEL Kinship Matrix",
+  "GWAS表型准备": "GWAS Phenotype Preparation",
+  "GWAS协变量准备": "GWAS Covariate Preparation",
+  "GWAS TASSEL MLM/CMLM": "GWAS TASSEL MLM/CMLM",
+  "GWAS MLM/CMLM绘图": "GWAS MLM/CMLM Plotting",
+  "GWAS EMMAX": "GWAS EMMAX",
+  "GWAS EMMAX绘图": "GWAS EMMAX Plotting",
+  "GWAS TASSEL GLM": "GWAS TASSEL GLM",
+  "群体PCA分析": "Population PCA Analysis",
+  "群体系统树分析": "Population Phylogenetic Tree",
+  "群体结构分析": "Population Structure Analysis",
+  "群体LD衰减分析": "Population LD Decay Analysis",
+  "SNP密度分析": "SNP Density Analysis",
+  "RNA-seq DE": "RNA-seq DE",
+  "组合图": "Combo Plot",
+  "参考亚家族分类": "Reference Subfamily Classification",
+  "亚家族标签确认": "Subfamily Label Confirmation",
+  "PPI/KaKs 参考数据": "PPI/KaKs Reference Data",
+  "PPI/KaKs 兜底确认": "PPI/KaKs Fallback Confirmation",
+  "亚家族确认输出": "Subfamily Confirmation Output",
+  "MSA区间重绘": "MSA Region Replot",
+  "MSA区间重绘已提交": "MSA Replot submitted",
+  "MSA区间重绘完成": "MSA Replot completed",
+  "MSA区间重绘失败": "MSA Replot failed",
+  "启动子自定义绘图已提交": "Promoter custom plot submitted",
+  "启动子自定义绘图完成": "Promoter custom plot completed",
+  "启动子自定义绘图失败": "Promoter custom plot failed",
+  "种内共线性 Circos 重绘已提交": "Intraspecies Circos replot submitted",
+  "种内共线性 Circos 重绘完成，已切换到新图": "Intraspecies Circos replot completed, switched to new figure",
+  "种内共线性 Circos 重绘失败": "Intraspecies Circos replot failed",
+  "亚家族确认 - Reference Classification": "Subfamily Confirmation - Reference Classification",
+  "这会停止当前 profile 的 Celery worker 并重新启动。若有任务正在运行，后端会默认拒绝执行。": "This will stop and restart the Celery worker for the current profile. Running tasks will be rejected by default.",
+  "Celery 已重启并完成队列检查": "Celery restarted and queue check complete",
+  "Celery 重启未完成": "Celery restart incomplete",
+  // ---- Card status_label (8) ----
+  "失败": "Failed",
+  "运行中": "Running",
+  "待确认": "Pending",
+  "待执行": "Pending",
+  "待核对": "Review",
+  "待授权": "Authorize",
+  "已取消": "Cancelled",
+  // ---- Card title (10) ----
+  "拓展分析候选核对": "Extension Candidate Review",
+  "拓展分析说明生成失败": "Description Generation Failed",
+  "正在生成拓展分析": "Generating Extension",
+  "运行结果核对": "Result Review",
+  "输入文件和参数": "Input Files & Parameters",
+  "正在恢复拓展分析状态": "Restoring Extension State",
+  // ---- Card purpose/summary (15) ----
+  "原始命令、测试数据和执行事实已经保留；这里只重新生成说明，不会重新运行原始命令。": "Raw command, test data and execution facts are preserved; only regenerate description without re-running.",
+  "可以直接重试，或取消后检查错误原因。": "Retry or cancel and check error cause.",
+  "这张卡用于核对系统从真实执行中锁定的命令、输入、参数和输出。": "Review system-locked command, inputs, parameters and outputs from real execution.",
+  "系统正在确定执行入口，并运行上传的代码和测试数据，提取不可修改的执行事实。": "Determining entry point, running uploaded code/test data, extracting immutable execution facts.",
+  "原始包可复现检查中…": "Checking reproducibility of original package…",
+  "这张卡用于用同一份测试数据执行已确认的工具，并生成可核对的结果。": "Run confirmed tool with same test data; generate verifiable results.",
+  "确认无误后点击“开始验证”。": "Click \"Start Validation\" after confirming.",
+  "这张卡用于查看实际命令、输出与日志；核对无误后才允许授权发布。": "View actual command, output and logs; publication requires review confirmation.",
+  "这张卡用于确认将已经验证的工具发布为可复用的拓展分析。": "Confirm publishing validated tool as reusable extension.",
+  "确认后才会发布；不会自动发布。": "Publication requires explicit confirmation; not automatic.",
+  "这张卡用于确认本次新数据运行的输入文件和原始命令已有参数；确认后立即执行，不会再次弹出确认。": "Confirm input files and parameters for this run; executes immediately without re-confirmation.",
+  "这张卡展示本次新数据运行生成的文件、标准输出和日志。": "Files, stdout and logs generated by this run.",
+  "页面已恢复；系统继续轮询同一个操作，不会重复提交。": "Page restored; continuing to poll the same operation without re-submitting.",
+  "请在页面下方的主对话输入框说明需要修订的内容，然后发送。命令和真实执行事实不会被修改。": "Describe revisions in the main chat input below. Commands and execution facts are immutable.",
+  "已取消；没有提交后续执行。": "Cancelled; no further execution submitted.",
+  // ---- Card rows labels (13) ----
+  "实际命令": "Actual Command",
+  "输出文件": "Output Files",
+  "标准输出": "Standard Output",
+  "见执行日志": "See execution log",
+  "功能": "Function",
+  "输入文件": "Input Files",
+  "可改参数": "Configurable Parameters",
+  "标准输出（直接展示）": "Stdout (direct display)",
+  "（无文件输出）": "(No file output)",
+  "（无）": "(None)",
+  // ---- Card error/misc (25+) ----
+  "操作失败": "Operation failed",
+  "读取拓展分析说明失败": "Failed to read extension help",
+  "读取拓展分析状态失败": "Failed to read extension status",
+  "阶段：": "Phase: ",
+  "原因：": "Cause: ",
+  "无法读取操作状态": "Cannot read operation status",
+  "提交失败": "Submission failed",
+  "原始包可复现检查提交失败": "Reproducibility check submission failed",
+  "阶段：upload": "Phase: Upload",
+  "已恢复修订；请核对最新候选。": "Revision restored; please review latest candidate.",
+  "✓ 语义说明已重新生成；执行命令未重新运行。": "✓ Description regenerated; command not re-run.",
+  "✓ 执行入口已成功复现；候选信息已生成。": "✓ Entry point reproduced successfully; candidate info generated.",
+  "✓ 验证已完成；结果核对卡已生成。": "✓ Validation complete; result review card generated.",
+  "✓ 已核对运行结果。": "✓ Results reviewed.",
+  "✓ 已显式授权并发布。": "✓ Explicitly authorized and published.",
+  "✓ 已确认输入与参数；本次运行已完成。": "✓ Inputs and parameters confirmed; run completed.",
+  "✓ 已恢复并完成；候选核对卡已生成。": "✓ Restored and completed; candidate review card generated.",
+  "✓ 已恢复确认；可以开始验证。": "✓ Confirmation restored; ready to validate.",
+  "✓ 已恢复验证；结果核对卡已生成。": "✓ Validation restored; result review card generated.",
+  "✓ 已恢复结果核对；可以授权发布。": "✓ Review restored; ready to publish.",
+  "✓ 已恢复并发布。": "✓ Restored and published.",
+  "✓ 已恢复运行；本次结果卡已生成。": "✓ Run restored; result card generated.",
+  "✓ 已确认候选｜未作修订": "✓ Candidate confirmed | no revision",
+  "✓ 已确认候选｜已修订：": "✓ Candidate confirmed | revised: ",
+  "文件上传成功，已开始生成并验收": "Upload successful; generation and validation started",
+  "请填写名称并提供测试数据文件或服务器路径；原始命令和原始代码至少提供一项。": "Please provide name and test data files or server path; at least command or code is required.",
+  "请写明修订意见。": "Please specify revision feedback.",
+  "读取验证结果失败": "Failed to read validation results",
+  "读取运行说明失败": "Failed to read run help",
+  "请先选择实际运行 Job。": "Please select a job first.",
+  "请填写每个输入文件。": "Please fill in each input file.",
+  "已修订：": "Revised: ",
+  "正在恢复 ": "Restoring ",
+  "…": "…",
+  " 提交失败": " submission failed",
+  " 原始包可复现检查提交失败": " Reproducibility check submission failed",
+  "执行入口已在隔离环境运行。请核对系统记录的事实与说明；确认后才会验证。": "Entry point ran in isolated environment. Review system-recorded facts and descriptions; validation proceeds after confirmation.",
+  // ---- Agent message: status/confirm/cancel (10) ----
+  "config.xlsx 已确认，继续当前目标：": "config.xlsx confirmed; continuing current goal: ",
+  "✅ 已确认使用拟南芥作为 PPI/KaKs 的显式兜底参考；该选择只写入 Ortholog/PPI/KaKs 专用参数，其它常规阶段不受影响。": "✅ Arabidopsis confirmed as PPI/KaKs explicit fallback reference; affects only Ortholog/PPI/KaKs parameters.",
+  "✅ 已确认自动注释，正在重新提交": "✅ Auto-annotation confirmed; re-submitting",
+  "✅ 已选择项目参考物种": "✅ Project reference species selected",
+  "已停止等待，本次未继续获取回答。": "Stopped waiting; no further response.",
+  "抱歉，我没有理解你的意思。请尝试更详细地描述。": "Sorry, I didn't understand. Please try describing in more detail.",
+  "抱歉，服务暂时不可用，请稍后重试。": "Sorry, service temporarily unavailable. Please try again later.",
+  // ---- Agent message: pause/block/skip (13) ----
+  "⏸️ PPI/KaKs 前置数据检查未通过，已暂停批量执行；请补齐参考注释包或确认参考物种后重新执行。": "⏸️ PPI/KaKs prerequisite check failed; batch paused. Add reference annotation or confirm reference species.",
+  "⏸️ identify 已完成，但方法学链路需要先确认基因家族成员表；保存成员确认后再继续 phylogeny。": "⏸️ identify completed; methodology chain requires confirmed gene family members before continuing phylogeny.",
+  "⏸️ 已打开 identify 成员初筛确认表；保存后会生成 curated_members，再继续亚家族证据阶段。": "⏸️ identify member confirmation table opened; save to generate curated_members, then continue subfamily evidence stages.",
+  "⏸️ 当前已有前台批次执行中；WGS 规划已保留，请在当前批次结束后重新执行。": "⏸️ Foreground batch in progress; WGS plan saved. Re-run after current batch finishes.",
+  "，其它阶段继续执行。": ", other stages continue.",
+  "⏸️ 已按选择先暂停 Ortholog/PPI/KaKs，其它阶段继续执行。": "⏸️ Paused Ortholog/PPI/KaKs as selected; other stages continue.",
+  "⏹️ 后端预检失败，当前批次已停止；请重新拉取规划后继续。": "⏹️ Backend pre-check failed; batch stopped. Re-pull plan to continue.",
+  // ---- Agent message: replan/continue (5) ----
+  "🔁 亚家族标签已确认，正在重新规划确认后常规基因家族分析...": "🔁 Subfamily labels confirmed; re-planning post-confirmation gene family analysis...",
+  "确认后常规分析没有新的可执行阶段。": "No new executable stages after confirmation.",
+  // ---- Agent message: submit/execute (7) ----
+  "⚙️ 提交阶段: ": "⚙️ Submitting stage: ",
+  "按 per_species 拆分 ": "Split by per_species ",
+  // ---- Agent message: complete/success (7) ----
+  "✅ PPI/KaKs 前置数据检查通过：当前批次已有 Ortholog，将先生成同源映射再继续下游分析。": "✅ PPI/KaKs check passed: Ortholog exists in current batch; generating mapping first.",
+  // ---- Agent message: download/background (5) ----
+  "⬇️ 正在提交推荐近源参考物种注释包后台下载；Ortholog/PPI/KaKs 将等待参考数据，其它阶段继续执行。": "⬇️ Submitting reference annotation download; Ortholog/PPI/KaKs wait for data, other stages continue.",
+  // ---- Agent message: autofill (2) ----
+  "已根据上传文件生成 samples.yaml 草稿。请在弹窗中核对 run、sample_id、condition、replicate、链特异性和分组角色，确认后再写入正式 samples.yaml。": "samples.yaml draft generated. Review run, sample_id, condition, replicate, strand specificity and group roles in the dialog, then confirm to write.",
+  "已根据上传文件生成 config.xlsx 草稿。请在弹窗中核对物种代码、文件类型、物种角色和 Gene_Family，确认后再写入正式 config.xlsx。": "config.xlsx draft generated. Review species code, file type, species role and Gene_Family in the dialog, then confirm to write.",
+  // ---- Agent message: methodology (4) ----
+  "已保存方法学方案到当前 job。下一步请上传/填写目标物种和参考物种数据，并确认 config.xlsx 后再执行方法学流程。": "Methodology plan saved. Next: upload target/reference species data, confirm config.xlsx, then execute.",
+  "❌ 方法学执行规划失败：": "❌ Methodology execution planning failed: ",
+  // ---- Agent message: replot/revision (4 to fix) ----
+  "进化树图重绘已提交：": "Phylogeny tree redraw submitted: ",
+  "进化树图重绘失败：": "Phylogeny tree redraw failed: ",
+  "Combo 图修订已提交：": "Combo plot revision submitted: ",
+  "Combo 图修订失败：": "Combo plot revision failed: ",
+  // ---- Agent message: failure/interrupt (10) ----
+  "无法执行分析：": "Cannot execute analysis: ",
+  "服务暂时不可用：": "Service temporarily unavailable: ",
+  "⚠️ 批量执行完成：成功 ": "⚠️ Batch complete: success ",
+  "⚠️ 以下阶段因依赖失败被跳过: ": "⚠️ Stages skipped due to dependency failure: ",
+  "⚠️ 以下阶段等待上游暂停项处理后继续: ": "⚠️ Stages waiting for upstream paused items: ",
+  "未生成可执行阶段。未检测到当前 scope 的完成结果时，不再把空计划解释为已完成；请检查阶段名称或前置依赖。": "No executable stages generated. Empty plan is not interpreted as completed. Check stage names or dependencies.",
+  // ---- Agent message: extension analysis (6) ----
+  "拓展分析阶段：": "Extension stage: ",
+  "拓展分析阶段失败：": "Extension stage failed: ",
+  "已执行拓展分析阶段：": "Extension stage executed: ",
+  "拓展分析阶段执行失败：": "Extension stage execution failed: ",
+  "拓展分析帮助暂时不可用：": "Extension help temporarily unavailable: ",
+  "上传可执行代码、测试数据和原始命令，核对候选后验证并授权发布；发布后可指定输入和原命令已有参数运行。": "Upload executable code, test data and original command; review candidate, validate, then authorize publishing.",
+  // ---- Agent message: other (5) ----
+  "📌 当前已有前台批次执行中，新任务已转为后台执行：": "📌 Foreground batch running; new task moved to background: ",
+  "已识别到 RNA-seq/FASTQ 和参考资源": "RNA-seq/FASTQ and reference resources detected",
+  "已把目录中的可识别参考资源导入当前 job": "Reference resources imported into current job",
+  "我识别到你是在提供数据路径，但该路径下没有发现": "Data path detected but no usable files found",
+  "已选择：": "Selected: ",
+  // ---- Agent message: remaining dynamic patterns ----
+  "❌ 批量执行已取消": "❌ Batch execution cancelled",
+  "⏭️ 跳过阶段: ": "⏭️ Skipped stage: ",
+  "🚀 准备提交依赖驱动执行，共 ": "🚀 Preparing dependency-driven execution: ",
+  "🎉 批量执行完成！成功: ": "🎉 Batch complete! Success: ",
+  "，失败 ": ", failed ",
+  "跳过阶段: ": "Skipped stage: ",
+  " 暂停：需要先准备参考物种 PEP/CDS 或 Ortholog 映射。": " paused: prepare reference species PEP/CDS or Ortholog mapping first.",
+  " 已暂停依赖 confirmed_subfamily.tsv 的后续阶段；请在亚家族确认表中保存后重新拉取规划继续执行。": " paused subsequent stages depending on confirmed_subfamily.tsv; save the subfamily confirmation table and re-pull the plan.",
+  "参考数据已就绪": "Reference data ready",
+  " 已恢复为待执行，将随当前批次继续调度。": " restored as pending; will continue with current batch.",
+  " 开始续跑 Ortholog/PPI/KaKs。": " resuming Ortholog/PPI/KaKs.",
+  " 重新提交（自动注释背景）": " re-submit (auto-annotation background)",
+  " 已提交后端，task_id=": " submitted to backend, task_id=",
+  " 准备执行 ": " preparing to execute ",
+  " 个阶段，请确认参数": " stages; please confirm parameters",
+  " 个阶段...": " stages...",
+  " 个分析阶段，请确认参数": " analysis stages; please confirm parameters",
+  "亚家族确认后已规划 ": "Post-subfamily-confirmation plan: ",
+  " 个常规分析阶段": " regular analysis stages",
+  "PPI/KaKs 前置数据检查通过：检测到已下载近源参考 ": "PPI/KaKs check passed: downloaded reference ",
+  "，将用该参考先生成 Ortholog mapping。": ", generating Ortholog mapping first.",
+  "PPI/KaKs 前置数据检查通过：检测到已有 Ortholog 映射（": "PPI/KaKs check passed: Ortholog mapping exists (",
+  "），将复用该映射执行 ": "), reusing mapping for ",
+  "），将直接执行 ": "), executing directly ",
+  "），已自动插入 Ortholog，将先生成 mapping 再执行 ": "), Ortholog auto-inserted; generating mapping then executing ",
+  "，已自动插入 Ortholog": ", Ortholog auto-inserted",
+  "推荐参考数据下载完成：": "Reference download complete: ",
+  "）。已写入 Ortholog/PPI/Ka/Ks 专用参考物种。": "). Written to Ortholog/PPI/Ka/Ks specific reference species.",
+  " 已按用户确认使用 bcftools 完成": " completed with bcftools as confirmed by user",
+  " 前端连接短暂中断，正在按后端产物状态恢复": " frontend connection interrupted; recovering from backend state",
+  "亚家族确认已完成，但后端当前 workflowPhase 为 ": "Subfamily confirmation completed; backend workflowPhase is ",
+  "，未自动打开常规分析。": "; no auto-analysis opened.",
+  "亚家族分类已完成预测，但确认表尚未保存；": "Subfamily classification predicted; confirmation table not yet saved; ",
+  "已调出": "Opened ",
+  "亚家族分类确认表。你可以在弹窗中修订 confirmed_subfamily，点击\"确认亚家族标签\"后保存。": " subfamily classification confirmation table. Revise confirmed_subfamily in the dialog, then click \"Confirm Subfamily Labels\" to save.",
+  "亚家族分类确认表没有打开：未找到可确认的分类结果。请先确认该物种已完成 reference_classification。": " subfamily classification confirmation table not opened: no confirmable results found. Confirm the species has completed reference_classification.",
+  "确认后常规分析没有新的可执行阶段。": "No new executable stages after confirmation.",
+  "⚠️ 检测到 ": "⚠️ Detected ",
+  "，将在确认后启动本地自动注释。": "; auto-annotation will start after confirmation.",
+  "🧬 自动背景注释与 ": "🧬 Auto-background annotation and ",
+  " 已提交后端，task_id=": " submitted to backend, task_id=",
+  "⚠️ 亚家族确认后自动续跑规划失败：": "⚠️ Post-subfamily-confirm auto-plan failed: ",
+  "; Ortholog/PPI/KaKs ": "; Ortholog/PPI/KaKs ",
+  "推荐参考数据下载已转入后台：": "Reference download moved to background: ",
+  "，task_id=": ", task_id=",
+  "后台批次完成：成功 ": "Background batch complete: success ",
+  "后台执行阶段: ": "Background stage: ",
+  "后台阶段 ": "Background stage ",
+  " 完成": " done",
+  "已保存方法学方案，但暂不执行：": "Methodology plan saved, not executing: ",
+  "。请在 config.xlsx 自动识别弹窗中核对/补齐数据，确认写入正式配置后再点击\"确认并执行\"。": ". Review/fill data in the config.xlsx auto-detect dialog, confirm to write then click \"Confirm & Execute\".",
+  "当前已有前台批次执行中；WGS 规划已保留，请在当前批次结束后重新执行。": "Foreground batch in progress; WGS plan saved. Re-run after current batch finishes.",
+  "拓展分析阶段执行失败：": "Extension stage execution failed: ",
+  "已把路径中的 ": "Soft-linked ",
+  " 个可识别文件软链接到当前 job": " recognizable files to current job",
+  " 个可识别文件软链接到当前 job，但没有发现需要自动填写 config.xlsx 或 samples.yaml 的文件类型；请检查目录或继续补充文件。": " recognizable files linked to current job; no files found for config.xlsx/samples.yaml auto-fill. Check directory or add more files.",
+  "前": " ago",
+  "秒前": "sec ago",
+  "分钟前": "min ago",
+  "小时前": "hr ago",
+  "天前": "d ago",
+  "待机": "Idle",
+  "分析中": "Analyzing",
+  "检索中": "Retrieving",
+  "反思中": "Reflecting",
+  "生成中": "Generating",
+  "完成": "Complete",
+  "错误": "Error",
+  "处理中...": "Processing...",
+  "发送": "Send",
+  "任务列表": "Tasks",
+  " 个常规分析阶段，请确认参数": " regular analysis stages; please confirm parameters",
+  " 亚家族分类已完成预测，但确认表尚未保存；后续会继续显示为待确认。": " subfamily classification predicted; confirmation table not yet saved; will show as pending.",
+  " 到后端；系统会先生成 ": " to backend; system will generate ",
+  " 前端连接短暂中断，正在按后端产物状态恢复，不会立即判定阶段失败": " frontend connection interrupted; recovering from backend artifact state. Stage will not be marked as failed.",
+  " 已有同一分析范围的有效结果，无需重复执行。": " already has valid results for the same analysis scope. No need to re-execute.",
+  " 的依赖尚未完成": " dependency not yet completed",
+  " 重新提交（自动注释背景）...": " re-submitting (auto-annotation background)...",
+  " 提交失败": " submission failed",
+  " 原始包可复现检查提交失败": " reproducibility check submission failed",
+  "Combo 图修订已提交：task_id=": "Combo plot revision submitted: task_id=",
+  "⏳ 推荐参考数据下载已转入后台：": "⏳ Reference download moved to background: ",
+  "⏸️ 以下阶段等待上游暂停项处理后继续: ": "⏸️ Stages waiting for upstream paused items: ",
+  "⚙️ 亚家族确认后已规划 ": "⚙️ Post-subfamily-confirm plan: ",
+  "⚙️ 准备执行 ": "⚙️ Preparing to execute ",
+  "⚙️ 后台执行阶段: ": "⚙️ Background stage: ",
+  "⚠️ 后台批次中以下阶段因依赖失败被跳过: ": "⚠️ Background batch stages skipped due to dependency failure: ",
+  "✅ PPI/KaKs 前置数据检查通过：检测到已下载近源参考 ": "✅ PPI/KaKs check passed: downloaded reference ",
+  "✅ PPI/KaKs 前置数据检查通过：检测到已有 Ortholog 映射（": "✅ PPI/KaKs check passed: Ortholog mapping found (",
+  "✅ 后台阶段 ": "✅ Background stage ",
+  "✅ 已确认自动注释，正在重新提交 ": "✅ Auto-annotation confirmed; re-submitting ",
+  "✅ 已选择项目参考物种 ": "✅ Selected project reference species: ",
+  "✅ 推荐参考数据下载完成：": "✅ Reference download complete: ",
+  "✅ 阶段 ": "✅ Stage ",
+  "☒ 已恢复并完成。": "☒ Restored and completed.",
+  "❌ 无法执行分析：": "❌ Cannot execute analysis: ",
+  "❌ 服务暂时不可用：": "❌ Service temporarily unavailable: ",
+  "。结果已写入当前 job workspace，输出：": ". Results written to current job workspace, output: ",
+  "。请在 config.xlsx 自动识别弹窗中核对/补齐数据，确认写入正式配置后再点击“确认并执行”。": ". Review/fill data in the config.xlsx auto-detect dialog; confirm to write then click \"Confirm & Execute\".",
+  "亚家族分类确认表。你可以在弹窗中修订 confirmed_subfamily，点击“确认亚家族标签”后保存。": "subfamily classification confirmation table. Revise confirmed_subfamily in the dialog, then click \"Confirm Subfamily Labels\" to save.",
+  "亚家族标签确认被跳过，已停止依赖 confirmed_subfamily.tsv 的后续阶段；如需普通无亚家族标签分析，请重新发起并明确说明\"不需要亚家族\"。": "Subfamily label confirmation skipped; subsequent stages depending on confirmed_subfamily.tsv have been stopped. For regular analysis without subfamily labels, please restart and explicitly state \"no subfamily\".",
+  "发送\"按已保存的方法学方案继续执行\"": "Send \"Continue with saved methodology plan\"",
+  "后台批次无法继续：": "Background batch cannot continue: ",
+  "已打开 config.xlsx 自动识别表。你可以直接修订单行，或用批量修订区筛选、预览、Apply。": "Opened config.xlsx auto-identification table. You can edit individual rows, or use the batch revision area to filter, preview, and Apply.",
+  "已打开 samples.yaml 自动识别表。你可以直接修订样本、分组、重复和对照组。": "Opened samples.yaml auto-identification table. You can revise samples, groups, replicates and controls directly.",
+  "已把目录中的可识别参考资源导入当前 job，并生成 config.xlsx 草稿。请在弹窗中核对和修订，点击“确认并写入正式配置”后再执行正式分析。": "Imported recognizable reference resources into current job and generated config.xlsx draft. Please review and revise in the dialog, then click \"Confirm & Write to Official Config\" before executing.",
+  "已识别到 RNA-seq FASTQ 和参考资源，优先打开 samples.yaml 草稿供你核对样本、分组和对照；config.xlsx 草稿也已生成。": "RNA-seq FASTQ and reference resources detected; opening samples.yaml draft for sample/group/control review. config.xlsx draft also generated.",
+  "已识别到 RNA-seq/FASTQ 数据并生成 samples.yaml 草稿。请在弹窗中核对 run、sample_id、condition、replicate、control_group 后再写入正式 samples.yaml。": "RNA-seq/FASTQ data detected and samples.yaml draft generated. Review run, sample_id, condition, replicate, control_group in the dialog, then confirm to write.",
+  "已重新识别并生成 config.xlsx 草稿。请在弹窗中核对并修订，确认后才会写入正式 config.xlsx。": "Re-identified and generated config.xlsx draft. Review and revise in the dialog; writes to official config.xlsx only after confirmation.",
+  "已重新识别并生成 samples.yaml 草稿。请确认 run、sample_id、condition、replicate 和 control_group 后再写入正式 samples.yaml。": "Re-identified and generated samples.yaml draft. Confirm run, sample_id, condition, replicate, and control_group before writing to official samples.yaml.",
+  "没有找到可展示的结果。": "No displayable results found.",
+  "点击\"浏览\"上传文献": "Click \"Browse\" to upload paper",
+  "点击\"知识图谱\"查看关系": "Click \"Knowledge Graph\" to view relations",
+  "请先核对 config.xlsx 草稿。": "Please review the config.xlsx draft first.",
+  "请补充你的需求。": "Please clarify your requirements.",
+  "进化树图重绘已提交：task_id=": "Phylogeny tree redraw submitted: task_id=",
+  "选择输入框上方\"Analysis\"": "Select \"Analysis\" above the input box",
+  "选择输入框上方\"RAG\"": "Select \"RAG\" above the input box",
+  "，再继续富集分析。": ", then continue enrichment analysis.",
+  "，将用该参考先生成 Ortholog mapping。": ", will generate Ortholog mapping from this reference first.",
+  "；Ortholog/PPI/KaKs 已恢复为待执行，将随当前批次继续调度。": "; Ortholog/PPI/KaKs restored as pending; will continue with current batch.",
+  "，将按该参考检查或生成 Ortholog，并执行 ": "; will check or generate Ortholog from this reference, and execute ",
+  "；开始续跑 Ortholog/PPI/KaKs。": "; resuming Ortholog/PPI/KaKs.",
+  "），将复用该映射执行 ": "), will reuse this mapping and execute ",
+  "），将直接执行 ": "), will execute directly ",
+  "，已自动插入 Ortholog，将先生成 mapping 再执行 ": ", Ortholog auto-inserted; generating mapping first then executing ",
+  "）。已写入 Ortholog/PPI/Ka/Ks 专用参考物种。": "). Written to Ortholog/PPI/Ka/Ks specific reference species.",
+  " 已按用户确认使用 bcftools 完成": " completed with bcftools as user confirmed",
+  " 已暂停依赖 confirmed_subfamily.tsv 的后续阶段；请在亚家族确认表中保存后重新拉取规划继续执行。": " has paused subsequent stages depending on confirmed_subfamily.tsv; save the subfamily confirmation table and re-pull the plan.",
+  " 暂停：需要先准备参考物种 PEP/CDS 或 Ortholog 映射。": "paused: prepare reference species PEP/CDS or Ortholog mapping first.",
+  " 个分析阶段，请确认参数": " analysis stages, please confirm parameters",
+  "📦 后台批次完成：成功 ": "📦 Background batch complete: success ",
+  "🔄 阶段 ": "🔄 Stage ",
+  "🚀 阶段 ": "🚀 Stage ",
+  "config.xlsx 已确认，继续当前目标：": "config.xlsx confirmed; continuing current goal: ",
+  "按 per_species 拆分 ": "Split by per_species ",
+  "⚠️ 检测到 ": "⚠️ Detected ",
+  "，将在确认后启动本地自动注释。": "; auto-annotation will start after confirmation.",
+  "。已写入 Ortholog/PPI/Ka/Ks 专用参考物种。": ". Written to Ortholog/PPI/Ka/Ks specific reference species.",
+  "🚀 准备提交依赖驱动执行，共 ": "🚀 Preparing dependency-driven execution: ",
+  " 个阶段...": " stages...",
+  "⚠️ 批量执行完成：成功 ": "⚠️ Batch complete: success ",
+  "，失败 ": ", failed ",
+  "🎉 批量执行完成！成功: ": "🎉 Batch complete! Success: ",
+  "⏭️ 跳过阶段: ": "⏭️ Skipped stage: ",
+  "⚙️ 提交阶段: ": "⚙️ Submitting stage: ",
+  "⏭️ ": "⏭️ ",
+  " 已提交后端，task_id=": " submitted to backend, task_id=",
+  " 完成": " complete",
+  "，task_id=": ", task_id=",
+  "⚠️ 亚家族确认后自动续跑规划失败：": "⚠️ Post-subfamily-confirm auto-plan failed: ",
+  "亚家族确认已完成，但后端当前 workflowPhase 为 ": "Subfamily confirmation complete; backend workflowPhase is ",
+  "，未自动打开常规分析。": "; no auto-analysis opened.",
+  "已调出": "Opened: ",
+  "亚家族分类确认表没有打开：未找到可确认的分类结果。请先确认该物种已完成 reference_classification。": "Subfamily classification confirmation table not opened: no confirmable results found. Confirm the species has completed reference_classification.",
+  "拓展分析帮助暂时不可用：": "Extension help temporarily unavailable: ",
+  "已保存方法学方案，但暂不执行：": "Methodology plan saved; not executing: ",
+  "🧬 自动背景注释与 ": "🧬 Auto-background annotation and ",
+  "参考数据已就绪": "Reference data ready",
+  "Combo 图修订失败：": "Combo plot revision failed: ",
+  "进化树图重绘失败：": "Phylogeny tree redraw failed: ",
+  "，其它阶段继续执行。": ", other stages continue.",
+  " 个阶段，请确认参数": " stages, please confirm parameters",
+  "—": "—",
+  "✓ 已恢复并完成。": "✓ Restored and completed.",
+  "后台批次完成：成功 ": "Background batch complete: success ",
+  "后台执行阶段: ": "Background stage: ",
+  "拓展分析阶段执行失败：": "Extension stage execution failed: ",
+  "拓展分析阶段：": "Extension stage: ",
+  "已执行拓展分析阶段：": "Extension stage executed: ",
+  "拓展分析阶段失败：": "Extension stage failed: ",
+  "MSA区间重绘已提交": "MSA replot submitted",
+  "MSA区间重绘失败": "MSA replot failed",
+  "已选择：": "Selected: ",
+  "📌 当前已有前台批次执行中，新任务已转为后台执行：": "📌 Foreground batch running; new task moved to background: ",
+  "，已自动插入 Ortholog，将先生成 mapping 再执行": ", Ortholog auto-inserted; generating mapping first then executing",
+  "推荐参考数据下载完成：": "Reference download complete: ",
+  "ABORTED": "ABORTED",
+  "的依赖尚未完成": "dependency not yet completed",
+  "亚家族标签确认被跳过，已停止依赖 confirmed_subfamily.tsv 的后续阶段；如需普通无亚家族标签分析，请重新发起并明确说明": "Subfamily label confirmation skipped; stages depending on confirmed_subfamily.tsv stopped. For regular analysis without subfamily labels, restart and explicitly state: ",
+  "不需要亚家族": "no subfamily",
+  "。": ".",
+  "，将复用该映射执行 ": ", will reuse this mapping and execute ",
+  "；将按该参考检查或生成 Ortholog，并执行 ": "; will check or generate Ortholog from this reference, and execute ",
+  "⚠️ 以下阶段因依赖失败被跳过: ": "⚠️ Stages skipped due to dependency failure: ",
+  "⚠️ 以下阶段等待上游暂停项处理后继续: ": "⚠️ Stages waiting for upstream paused items: ",
+  "启动子自定义绘图已提交": "Promoter custom plot submitted",
+  "启动子自定义绘图失败": "Promoter custom plot failed",
+  "种内共线性 Circos 重绘已提交": "Intraspecies Circos replot submitted",
+  "种内共线性 Circos 重绘失败": "Intraspecies Circos replot failed",
+  " 重新提交（自动注释背景）": " re-submit (auto-annotation background)",
+  " 已暂停依赖 confirmed_subfamily.tsv": " has paused stages depending on confirmed_subfamily.tsv",
+  "；": ";",
+  " 个分析阶段": " analysis stages",
+  " 个阶段": " stages",
+
+  // ===== 阶段 7：ElMessage/ElMessageBox i18n keys =====
+  // 注意：key 为中文原文（zh 模式直接返回 key），value 为 English
+  // tp() 类 key 需要同时注册变量 placeholder 版本和 static 文案
+
+  // ElMessageBox
+  "删除会话": "Delete Session",
+  "取消 PPI": "Cancel PPI",
+  "自动注释并继续": "Auto-annotate & Continue",
+  "我先去上传背景文件": "Upload Background Files First",
+  "确认设计并写入": "Confirm Design & Write",
+  "返回修改": "Return to Modify",
+  "确认第一候选": "Confirm First Candidate",
+  "使用 bcftools 重试": "Retry with bcftools",
+  "GATK 执行失败": "GATK Execution Failed",
+  "确认 DE 分析设计": "Confirm DE Analysis Design",
+  "确认 PPI 的 STRING 近源参考物种": "Confirm STRING Close Reference Species for PPI",
+
+  // autoAnnotation 标题
+  "GO/KEGG 富集背景缺失": "GO/KEGG Enrichment Background Missing",
+  "KEGG 富集背景缺失": "KEGG Enrichment Background Missing",
+  "GO 富集背景缺失": "GO Enrichment Background Missing",
+  "RNA-seq 富集背景缺失": "RNA-seq Enrichment Background Missing",
+
+  // 会话管理
+  "示例下载功能当前不可用": "Example download is currently unavailable",
+  "已Clear当前会话": "Current session cleared",
+
+  // autoFill
+  "请至少填写一个要预览修改的字段": "Please fill at least one field to preview modifications",
+  "没有匹配的草稿行": "No matching draft rows",
+  "请先预览要应用的修改": "Please preview modifications before applying",
+  "请先生成或加载草稿": "Please generate or load a draft first",
+  "生成草稿失败": "Failed to generate draft",
+  "确认失败": "Confirmation failed",
+  "已写入正式 config.xlsx": "Formal config.xlsx written",
+  "已写入正式 samples.yaml": "Formal samples.yaml written",
+  "请输入文件路径": "Please enter file path",
+  "加载失败": "Load failed",
+
+  // RAG 解读
+  "请求解读服务失败": "Interpretation service request failed",
+  "RAG解读已刷新": "RAG interpretation refreshed",
+  "RAG解读刷新失败": "RAG interpretation refresh failed",
+  "下载报告失败": "Failed to download report",
+  "JSON数据已加载到控制台": "JSON data loaded to console",
+  "解读内容已复制": "Interpretation content copied",
+  "复制失败": "Copy failed",
+
+  // 知识图谱
+  "知识图谱加载完成": "Knowledge graph loaded",
+  "未找到匹配的节点": "No matching node found",
+  "视图已重置": "View reset",
+  "请先通过问答加载子图，或点击「重新加载」": "Please load subgraph via Q&A first, or click Reload",
+
+  // PPI/STRING
+  "当前没有可用 job_id": "No job_id currently available",
+  "该 ID 已存在": "This ID already exists",
+  "至少保留一个 gene ID": "Keep at least one gene ID",
+
+  // Working Directory
+  "切换失败": "Switch failed",
+
+  // 图片/可视化
+  "请先选择一张图片": "Please select an image first",
+  "无法确定启动子输出前缀。": "Cannot determine promoter output prefix.",
+  "请至少选择一个启动子元件或类别。": "Please select at least one promoter element or category.",
+  "请至少选择一种图类型。": "Please select at least one plot type.",
+  "无法确定进化树输出标签，请先选择树图或填写输出标签。": "Cannot determine phylogeny tree output label. Please select a tree image or fill in the output label first.",
+  "进化树图重绘完成，已切换到新图": "Phylogeny tree replot complete; switched to new image",
+  "无法确定种内共线性输出标签，请先选择 Circos 图。": "Cannot determine intraspecies synteny output label. Please select a Circos image first.",
+  "无法确定 combo 输出标签，请先选择 combo 图。": "Cannot determine combo output label. Please select a combo image first.",
+  "Combo 修订至少选择两个类型。": "Combo revision requires at least two panel types.",
+  "Combo 图修订完成，已切换到新图": "Combo plot revision complete; switched to new image",
+  "无法确定MSA输出标签，请先选择对应图片或填写输出标签。": "Cannot determine MSA output label. Please select corresponding image or fill in output label.",
+  "请填写MSA alignment 起始列和终止列，或选择其他模式。": "Please fill in MSA alignment start/end columns or choose another mode.",
+  "请填写参考基因及其起止残基编号。": "Please fill in reference gene and its start/end residue numbers.",
+
+  // GWAS/诊断
+  "请填写 GWAS 物种简称（用于结果前缀）": "Please fill in GWAS species abbreviation (for result prefix)",
+  "请先选择有效 VCF": "Please select a valid VCF first",
+  "请先SelectAnalysis Stage": "Please select an analysis stage first",
+  "诊断未发现明确问题，请查看阶段错误日志。": "Diagnosis found no clear issues. Please check stage error logs.",
+  "诊断仅提示信息，无需确认修复。": "Diagnosis is informational only; no fix confirmation needed.",
+  "✅ 已应用修复方案，继续执行...": "✅ Fix applied; continuing execution...",
+  "❌ 应用修复失败": "❌ Fix application failed",
+  "❌ 网络错误，请重试": "❌ Network error; please retry",
+  "⏸ 已跳过修复；该问题会阻断下游，后续阶段已暂停。": "⏸ Fix skipped; issue blocks downstream; subsequent stages paused.",
+  "已关闭诊断提示。": "Diagnosis prompt closed.",
+
+  // 任务/消息
+  "当前消息未成功写入会话历史，刷新后可能丢失": "Message not persisted to session history; may be lost on refresh",
+  "路径已复制": "Path copied",
+
+  // 拓展分析
+  "当前筛选没有成员，未重建 final_members。": "No members matching current filter; final_members not rebuilt.",
+  "当前选择为按物种分别分析，但后端没有返回完整的分物种亚家族结果；将尝试打开可用的联合结果。": "Per-species analysis selected but backend returned incomplete results; attempting to open available joint results.",
+  "亚家族预测已完成，但未能自动打开确认表；请发送\"调出亚家族确认表\"手动打开。": "Subfamily prediction completed but confirmation table could not open automatically; please send \"show subfamily review table\" manually.",
+  "没有可确认的亚家族标签。": "No subfamily labels to confirm.",
+  "已请求停止当前拓展分析；状态仍会继续轮询直到进程退出。": "Stop requested for current extension; polling will continue until process exits.",
+  "No running task to stop": "No running task to stop",
+  "Task(s) stopped (backend + Celery when available)": "Task(s) stopped (backend + Celery when available)",
+  "File uploaded successfully": "File uploaded successfully",
+  "未检测到可用文件": "No usable files detected",
+  "请先生成拓展分析阶段。": "Please generate the extension stage first.",
+  "发布前验收已通过": "Pre-publish validation passed",
+  "请填写名称，并上传原始代码。": "Please provide name and upload source code.",
+  "拓展分析阶段已生成并发布": "Extension stage generated and published",
+  "确认已保存": "Confirmation saved",
+  "验证成功": "Validation successful",
+  "验证未通过": "Validation not passed",
+  "拓展分析阶段已发布": "Extension stage published",
+  "拓展分析阶段执行完成": "Extension stage execution complete",
+  "拓展分析阶段执行未成功，请查看结果日志": "Extension stage execution incomplete; please check result logs",
+  "该阶段已在正式列表中": "This stage is already in the formal list",
+  "已加入正式阶段列表，并默认纳入本次执行": "Added to formal stage list and included in this execution",
+
+  // 方法学
+  "请先选择 PDF 或 MinerU content_list JSON 文献文件": "Please select a PDF or MinerU content_list JSON literature file first",
+  "方法学提取启动失败": "Methodology extraction failed to start",
+  "方法学 sandbox 示例读取失败": "Failed to read methodology sandbox examples",
+  "请先选择或输入 methodology_dir": "Please select or enter methodology_dir first",
+  "方法学 sandbox 已构建": "Methodology sandbox built",
+  "请先构建临时库": "Please build the sandbox first",
+  "请输入方法学问题": "Please enter methodology question",
+  "方法学 sandbox 查询失败": "Methodology sandbox query failed",
+  "方法学方案已保存，可先准备数据，稍后再按该方案执行": "Methodology plan saved; you may prepare data first and execute later",
+  "方法学方案保存失败": "Failed to save methodology plan",
+  "当前已有批量任务在执行，请等待完成后再提交方法学流程": "A batch task is currently running; please wait for completion before submitting methodology workflow",
+  "请至少选择一个可执行的 bbav5 阶段": "Please select at least one executable bbav5 stage",
+  "亚家族参考物种未写入当前 job 配置，已生成方法学本地计划，请在批量确认页核对参考物种": "Subfamily reference species not written to current job config; local methodology plan generated. Please verify reference species on batch confirmation page.",
+  "方法学执行规划失败": "Methodology execution planning failed",
+
+  // Other
+  "请Input内容": "Please enter content",
+  "请先选择或创建 Working Directory": "Please select or create a Working Directory first",
+  "规划请求失败": "Plan request failed",
+  "重新绘图功能开发中...": "Replot feature is under development...",
+
+  // ===== tp() keys — 使用完整中文模板作为 key =====
+  "确定删除会话「{title}」吗？该操作会从列表中移除，但保留后台记录。": "Delete session \"{title}\"? This removes it from the list but keeps backend records.",
+  "切换到: {title}": "Switched to: {title}",
+  "预览 {rows} 行：{fields}。确认无误后点击 Apply。": "Preview {rows} rows: {fields}. Click Apply once confirmed.",
+  "已应用 {rows} 行修改。": "Applied {rows} row modifications.",
+  "已取消预览，表格已恢复到预览前。": "Preview cancelled, table restored.",
+  "链特异性为必填：还有 {count} 个样本未填写（{examples}）": "Library strandness required: {count} sample(s) unfilled ({examples})",
+  "设计提示：{error}；将保留原流程的兼容处理。": "Design tip: {error}; keeping original compatible handling.",
+  "已加载 {count} files": "Loaded {count} files",
+  "加载图谱失败: {message}": "Graph load failed: {message}",
+  "找到节点: {name}": "Node found: {name}",
+  "读取表格失败: {reason}": "Table read failed: {reason}",
+  "PPI Reset 已完成: {reset_id}": "PPI Reset complete: {reset_id}",
+  "STRING 链接文件内容：{preview}": "STRING link file content: {preview}",
+  "读取文件失败: {message}": "File read failed: {message}",
+  "已创建新Working Directory: {job_id}": "New Working Directory created: {job_id}",
+  "创建失败: {message}": "Creation failed: {message}",
+  "获取Working Directory列表失败: {message}": "Failed to get Working Directory list: {message}",
+  "已切换到Working Directory: {job_id}": "Switched to Working Directory: {job_id}",
+  "切换失败: {message}": "Switch failed: {message}",
+  "进化树图重绘失败：{message}": "Phylogeny tree replot failed: {message}",
+  "Combo 图修订失败：{message}": "Combo plot revision failed: {message}",
+  "任务: {message}\n状态: {status}\n进度: {progress}%": "Task: {message}\nStatus: {status}\nProgress: {progress}%",
+  "停止拓展分析失败：{message}": "Failed to stop extension: {message}",
+  "已选择 {count} 个原始代码文件": "{count} code file(s) selected",
+  "已选择 {count} 个测试数据文件": "{count} test data file(s) selected",
+  "已拖入 {count} 个文件，已自动填充拓展包": "{count} file(s) dropped; extension package auto-filled",
+  "❌ {message}": "❌ {message}",
+  "Upload complete. Latest Excel is now used as config.xlsx.{warningText}": "Upload complete. Latest Excel is now used as config.xlsx.{warningText}",
+  "{title}，需要自动注释": "{title}, Auto-annotation Required",
+
+  // String close species (tp with multi-line)
+  "目标物种不在 STRING 固定快照中。请确认是否采用排序第一的近源物种：": "Target species not in STRING fixed snapshot. Confirm using the top-ranked close species:",
+  "确认后才会下载参考 PEP、生成 Ortholog，并执行 PPI。": "Reference PEP will be downloaded, Ortholog generated, and PPI executed only after confirmation.",
+  "是否切换为\"自动注释 + 继续富集\"？": "Switch to \"Auto-annotation + Continue Enrichment\"?"
+};
+

gfa_core/extension_stage_lightweight/api.py

@@ -361,6 +361,7 @@ def execute_queued_operation(payload: dict[str, Any]) -> dict[str, Any]:
 
     settings = LightweightApiSettings.from_environment()
     action = str(payload["action"])
+    runtime_env = str(payload.get("runtime_env", "test4py311"))
     extension_id = _safe_component(str(payload["extension_id"]), name="extension_id")
     token = _safe_component(str(payload["operation_token"]), name="operation_token")
     _set_receipt(settings, token, state="running", status="running")
@@ -373,6 +374,7 @@ def execute_queued_operation(payload: dict[str, Any]) -> dict[str, Any]:
                 raw_code=stage / "raw_code",
                 raw_data=stage / "raw_test_data",
                 original_command=str(payload["original_command"]),
+                runtime_env=runtime_env,
                 operation_token=token,
             )
         elif action == "rebuild":
@@ -479,6 +481,7 @@ def create_lightweight_extension_router(
         raw_code_files: list[UploadFile] | None = File(None),
         raw_data_files: list[UploadFile] | None = File(None),
         raw_data_paths: str | None = Form(None),
+        runtime_env: str = Form("test4py311"),
     ) -> dict[str, Any]:
         temporary_stage: Path | None = None
         try:
@@ -555,6 +558,7 @@ def create_lightweight_extension_router(
                                 "action": "upload",
                                 "operation_token": token,
                                 "original_command": normalized_command,
+                                "runtime_env": runtime_env,
                             }
                         )
                     except BaseException as exc:

gfa_core/extension_stage_lightweight/service.py

@@ -109,6 +109,18 @@ class LightweightExtensionService:
     def get(self, extension_id: str) -> dict[str, Any]:
         return self._read_record(extension_id)
 
+    def _make_env(self, runtime_env: str) -> dict[str, str]:
+        """Build a process env that activates the selected conda environment."""
+
+        conda_prefix = f"/home/xzq/miniconda3/envs/{runtime_env}"
+        conda_bin = f"{conda_prefix}/bin"
+        env = dict(self.environment or {})
+        existing_path = env.get("PATH", os.environ.get("PATH", ""))
+        env["PATH"] = f"{conda_bin}:{existing_path}" if existing_path else conda_bin
+        env["CONDA_PREFIX"] = conda_prefix
+        env["CONDA_DEFAULT_ENV"] = runtime_env
+        return env
+
     def _deepseek_candidate(self, facts: dict[str, Any], raw_code: Path) -> tuple[dict[str, Any], dict[str, Any]]:
         return generate_semantic_candidate(
             execution_facts=facts, raw_code=raw_code, environ=self.environment
@@ -324,12 +336,13 @@ class LightweightExtensionService:
         build: dict[str, Any] | None = None
         failure: tuple[str, BaseException | str] | None = None
         try:
+            runtime_env = str(record.get("runtime_env") or "test4py311")
             evidence = run_raw_baseline(
                 command=record["original_command"],
                 raw_code=root / "raw_code",
                 raw_data=root / "raw_test_data",
                 run_dir=root / "baseline" / run_id,
-                env=self.environment,
+                env=self._make_env(runtime_env),
                 on_process_start=lambda identity: self._mark_process(
                     extension_id, token=token, phase="baseline", identity=identity
                 ),
@@ -365,6 +378,7 @@ class LightweightExtensionService:
         raw_code: Path,
         raw_data: Path,
         original_command: str,
+        runtime_env: str = "test4py311",
         operation_token: str,
     ) -> dict[str, Any]:
         raw_code, raw_data = Path(raw_code), Path(raw_data)
@@ -394,6 +408,7 @@ class LightweightExtensionService:
                     "schema_version": "lightweight-extension-service/v1",
                     "extension_id": extension_id,
                     "original_command": original_command,
+                    "runtime_env": runtime_env,
                     "status": "baseline_running",
                     "operations": {
                         operation_token: {
@@ -605,9 +620,10 @@ class LightweightExtensionService:
         rendered = render_runtime_package(
             record["runtime_binding_contract"], root / "raw_code", resolved, run_dir
         )
+        runtime_env = str(record.get("runtime_env") or "test4py311")
         result = replay_runtime_binding(
             rendered,
-            env=self.environment,
+            env=self._make_env(runtime_env),
             on_process_start=lambda identity: self._mark_process(
                 record["extension_id"], token=operation_token, phase=phase, identity=identity
             ),

gfa_core/gene_family_report.py

@@ -0,0 +1,383 @@
+"""Generate the formal report for one gene-family analysis plan."""
+
+from __future__ import annotations
+
+from dataclasses import dataclass
+from datetime import datetime
+import json
+from pathlib import Path
+import re
+from typing import Any, Iterable, Mapping
+
+from gfa_core.gene_family_report_data import (
+    ReportData,
+    ReportSection,
+    build_report_data,
+)
+from gfa_core.gene_family_report_html import (
+    REPORT_RENDERER_VERSION,
+    collect_stage_figures,
+    markdown_to_html,
+    render_stage_figures,
+)
+from gfa_core.mas.agents.interpretation_evidence import load_stage_manifests
+
+
+REPORT_STAGE = "gene_family_report"
+REPORT_TEMPLATE_VERSION = "gene_family_report_template_v1"
+EXCLUDED_PLAN_STAGES = {"summary", REPORT_STAGE, "subfamily_confirm"}
+TEMPLATE_PATH = Path(__file__).with_name("templates") / "gene_family_report_v1.md"
+CONNECTORS = {
+    "also": "此外，",
+    "however": "不过，",
+    "therefore": "因此，",
+}
+
+
+class NoSuccessfulGeneFamilyEvidence(RuntimeError):
+    """Raised when the requested plan has no successful result to report."""
+
+
+@dataclass(frozen=True)
+class GeneFamilyReportResult:
+    report_path: Path
+    manifest_path: Path
+    report_status: str
+    generation_mode: str
+    included_stages: tuple[str, ...]
+    reused: bool = False
+    html_path: Path | None = None
+
+
+def _normalized_stages(stages: Iterable[str] | None) -> list[str]:
+    result: list[str] = []
+    for raw in stages or ():
+        stage = str(raw).strip()
+        if stage and stage not in EXCLUDED_PLAN_STAGES and stage not in result:
+            result.append(stage)
+    return result
+
+
+def _parse_json_object(text: str) -> dict[str, Any]:
+    value = text.strip()
+    if value.startswith("```"):
+        value = re.sub(r"^```(?:json)?\s*", "", value, flags=re.IGNORECASE)
+        value = re.sub(r"\s*```$", "", value)
+    parsed = json.loads(value)
+    if not isinstance(parsed, dict):
+        raise ValueError("model output must be a JSON object")
+    return parsed
+
+
+def _api_payload(report_data: ReportData) -> dict[str, Any]:
+    return {
+        "sections": [
+            {
+                "section_id": section.section_id,
+                "sentences": [
+                    {"sentence_id": sentence.id, "text": sentence.text}
+                    for sentence in section.sentences
+                ],
+            }
+            for section in report_data.sections
+        ],
+        "allowed_connectors": list(CONNECTORS),
+    }
+
+
+def _generate_api_narrative(payload: dict, stages: list[str]) -> tuple[dict[str, Any], dict[str, Any]]:
+    from gfa_core.mas.config import config as mas_config
+    from gfa_core.mas.llm.base import LLMMessage
+    from gfa_core.mas.llm.manager import LLMManager
+
+    prompt = (
+        "仅编排以下不可改写句子。输出 JSON 字段只能是 sections；每节只能含 section_id、paragraphs，"
+        "每段只能含 sentence_ids、connector_ids。必须保留所有章节和句子且各出现一次，"
+        "句子不得跨章节；connector_ids 只能来自 allowed_connectors。计划阶段为："
+        f"{','.join(stages)}。\n"
+        + json.dumps(payload, ensure_ascii=False)
+    )
+    response = LLMManager(config=mas_config.load_llm_config()).call(
+        [
+            LLMMessage(role="system", content="你是证据约束的生物信息学报告编辑器，只返回 JSON。"),
+            LLMMessage(role="user", content=prompt),
+        ],
+        track_agent_name="GeneFamilyReport",
+    )
+    return _parse_json_object(response.content), {
+        "status": "generated",
+        "provider": response.provider,
+        "model": response.model,
+    }
+
+
+def _default_arrangement(report_data: ReportData) -> dict[str, Any]:
+    return {
+        "sections": [
+            {
+                "section_id": section.section_id,
+                "paragraphs": [
+                    {"sentence_ids": [sentence.id], "connector_ids": []}
+                    for sentence in section.sentences
+                ],
+            }
+            for section in report_data.sections
+        ]
+    }
+
+
+def _validate_arrangement(arrangement: Any, report_data: ReportData) -> None:
+    if not isinstance(arrangement, dict) or set(arrangement) != {"sections"}:
+        raise ValueError("API response has unknown or missing root fields")
+    sections = arrangement["sections"]
+    if not isinstance(sections, list):
+        raise ValueError("API sections must be a list")
+    expected = {section.section_id: {sentence.id for sentence in section.sentences} for section in report_data.sections}
+    seen_sections: set[str] = set()
+    for section in sections:
+        if not isinstance(section, dict) or set(section) != {"section_id", "paragraphs"}:
+            raise ValueError("API section has unknown or missing fields")
+        section_id = section["section_id"]
+        if not isinstance(section_id, str) or section_id not in expected or section_id in seen_sections:
+            raise ValueError("API changed or repeated a report section")
+        seen_sections.add(section_id)
+        paragraphs = section["paragraphs"]
+        if not isinstance(paragraphs, list):
+            raise ValueError("API paragraphs must be a list")
+        seen_sentences: list[str] = []
+        for paragraph in paragraphs:
+            if not isinstance(paragraph, dict) or set(paragraph) != {"sentence_ids", "connector_ids"}:
+                raise ValueError("API paragraph has unknown or missing fields")
+            sentence_ids = paragraph["sentence_ids"]
+            connector_ids = paragraph["connector_ids"]
+            if not isinstance(sentence_ids, list) or not sentence_ids or any(not isinstance(value, str) for value in sentence_ids):
+                raise ValueError("API sentence_ids must be a non-empty string list")
+            if not isinstance(connector_ids, list) or len(connector_ids) != len(sentence_ids) - 1:
+                raise ValueError("API connector count does not match its paragraph")
+            if any(not isinstance(value, str) or value not in CONNECTORS for value in connector_ids):
+                raise ValueError("API used an unknown connector")
+            if any(value not in expected[section_id] for value in sentence_ids):
+                raise ValueError("API used an unknown or cross-section sentence")
+            seen_sentences.extend(sentence_ids)
+        if len(seen_sentences) != len(set(seen_sentences)) or set(seen_sentences) != expected[section_id]:
+            raise ValueError("API omitted or repeated a required sentence")
+    if seen_sections != set(expected):
+        raise ValueError("API omitted a report section")
+
+
+def _template_blocks() -> list[tuple[str, str]]:
+    template = TEMPLATE_PATH.read_text(encoding="utf-8")
+    marker = re.compile(r"(?m)^<!-- GFA_SECTION ([a-z0-9_.]+) -->\s*$")
+    matches = list(marker.finditer(template))
+    if not matches:
+        raise ValueError(f"报告模板没有章节边界：{TEMPLATE_PATH}")
+    return [
+        (
+            match.group(1),
+            template[match.end(): matches[index + 1].start() if index + 1 < len(matches) else len(template)].strip(),
+        )
+        for index, match in enumerate(matches)
+    ]
+
+
+def _arranged_content(section: ReportSection, arranged: Mapping[str, Any]) -> str:
+    sentences = {sentence.id: sentence.text for sentence in section.sentences}
+    paragraphs: list[str] = []
+    for paragraph in arranged["paragraphs"]:
+        sentence_ids = paragraph["sentence_ids"]
+        text = sentences[sentence_ids[0]]
+        for connector_id, sentence_id in zip(paragraph["connector_ids"], sentence_ids[1:]):
+            text += f" {CONNECTORS[connector_id]}{sentences[sentence_id]}"
+        paragraphs.append(text)
+    return "\n\n".join(paragraphs)
+
+
+def _render_completion_table(report_data: ReportData) -> str:
+    supported_stages = {section.stage for section in report_data.sections if section.stage}
+    included = set(report_data.included)
+    status_labels = {"success": "成功", "partial_success": "部分完成", "failed": "失败", "missing": "未完成"}
+    lines = [
+        f"本次计划 {len(report_data.planned)} 个阶段，纳入解读 {len(report_data.included)} 个，未纳入 {len(report_data.planned) - len(report_data.included)} 个。",
+        "",
+        "| 分析阶段 | 运行状态 | 模板支持 | 纳入报告 |",
+        "|---|---|---|---|",
+    ]
+    for stage in report_data.planned:
+        status = str(report_data.statuses.get(stage, "missing"))
+        lines.append(
+            f"| `{stage}` | {status_labels.get(status, status)} | "
+            f"{'是' if stage in supported_stages else '否'} | {'是' if stage in included else '否'} |"
+        )
+    return "\n".join(lines)
+
+
+def _render_markdown(
+    report_data: ReportData,
+    *,
+    job_id: str,
+    generation_mode: str,
+    arrangement: Mapping[str, Any],
+    figures_by_stage: Mapping[str, list[dict[str, str]]],
+) -> str:
+    section_data = {section.section_id: section for section in report_data.sections}
+    arranged = {section["section_id"]: section for section in arrangement["sections"]}
+    rendered: list[str] = []
+    for section_id, block in _template_blocks():
+        if section_id not in section_data:
+            continue
+        content = (
+            _render_completion_table(report_data)
+            if section_id == "common.completeness"
+            else _arranged_content(section_data[section_id], arranged[section_id])
+        )
+        stage = section_id.removeprefix("stage.") if section_id.startswith("stage.") else ""
+        if stage and figures_by_stage.get(stage):
+            content += render_stage_figures(stage, figures_by_stage[stage])
+        rendered.append(block.replace("{{SECTION_CONTENT}}", content))
+    markdown = "\n\n".join(rendered)
+    fallback_notice = (
+        "> 生成模式：候补脚本。模型 API 不可用或输出越过语言编排边界；报告事实仍来自确定性结果数据。"
+        if generation_mode == "fallback_script"
+        else ""
+    )
+    replacements = {
+        "{{PREFIX}}": report_data.prefix,
+        "{{JOB_ID}}": job_id or "未指定",
+        "{{GENERATION_MODE}}": "API 编排" if generation_mode == "api" else "候补脚本",
+        "{{GENERATED_AT}}": datetime.now().isoformat(timespec="seconds"),
+        "{{FALLBACK_NOTICE}}": fallback_notice,
+    }
+    for placeholder, value in replacements.items():
+        markdown = markdown.replace(placeholder, value)
+    if "{{" in markdown or "【填写】" in markdown:
+        raise ValueError("报告模板仍含未替换占位符")
+    return markdown.rstrip() + "\n"
+
+
+def generate_gene_family_report(
+    workspace: Path | str,
+    prefix: str,
+    *,
+    expected_stages: Iterable[str] | None = None,
+    stage_statuses: Mapping[str, str] | None = None,
+    scope_id: str = "",
+    analysis_route: str = "",
+    member_set_id: str = "",
+    rnaseq_scope_id: str = "",
+    use_model: bool = True,
+    action: str = "generate",
+    job_id: str = "",
+) -> GeneFamilyReportResult:
+    if action not in {"generate", "open"}:
+        raise ValueError("action must be 'generate' or 'open'")
+    final_dir = Path(workspace).expanduser().resolve() / "final"
+    report_path = final_dir / f"{prefix}.gene_family_report.md"
+    html_path = final_dir / f"{prefix}.gene_family_report.html"
+    manifest_path = final_dir / f"{prefix}.gene_family_report.manifest.json"
+    if action == "open" and report_path.is_file() and manifest_path.is_file():
+        manifest = json.loads(manifest_path.read_text(encoding="utf-8"))
+        if not html_path.is_file():
+            html_path.write_text(markdown_to_html(report_path.read_text(encoding="utf-8"), report_path, f"{prefix} 基因家族分析结果解读报告"), encoding="utf-8")
+        manifest.setdefault("artifacts", {})["gene_family_report_html"] = {"path": f"final/{html_path.name}", "full_path": str(html_path), "type": "html"}
+        manifest_path.write_text(json.dumps(manifest, ensure_ascii=False, indent=2) + "\n", encoding="utf-8")
+        metrics = manifest.get("metrics") or {}
+        return GeneFamilyReportResult(
+            report_path,
+            manifest_path,
+            str(metrics.get("report_status") or "partial"),
+            str(metrics.get("generation_mode") or "fallback_script"),
+            tuple(metrics.get("included_stages") or []),
+            reused=True,
+            html_path=html_path,
+        )
+
+    planned = _normalized_stages(expected_stages)
+    if not planned:
+        raise ValueError("expected_stages is required when generating a gene-family report")
+    manifests = load_stage_manifests(
+        final_dir,
+        prefix,
+        scope_id=scope_id,
+        analysis_route=analysis_route,
+        member_set_id=member_set_id,
+        rnaseq_scope_id=rnaseq_scope_id,
+    )
+    statuses = {str(key): str(value) for key, value in (stage_statuses or {}).items()}
+    for stage in planned:
+        statuses.setdefault(stage, str((manifests.get(stage) or {}).get("status") or "missing"))
+    report_data = build_report_data(final_dir, prefix, planned, statuses, manifests)
+    if not report_data.included:
+        raise NoSuccessfulGeneFamilyEvidence("暂无可生成的基因家族分析结果")
+
+    arrangement = _default_arrangement(report_data)
+    generation_mode = "fallback_script"
+    model_metadata: dict[str, Any] = {"status": "disabled"}
+    if use_model:
+        try:
+            candidate, model_metadata = _generate_api_narrative(_api_payload(report_data), list(report_data.included))
+            _validate_arrangement(candidate, report_data)
+            arrangement = candidate
+            generation_mode = "api"
+        except Exception:
+            model_metadata = {
+                "status": "fallback",
+                "reason": "api_unavailable_or_boundary_violation",
+            }
+
+    figures_by_stage = collect_stage_figures(final_dir, report_data.included, manifests)
+    markdown = _render_markdown(
+        report_data,
+        job_id=job_id,
+        generation_mode=generation_mode,
+        arrangement=arrangement,
+        figures_by_stage=figures_by_stage,
+    )
+    final_dir.mkdir(parents=True, exist_ok=True)
+    report_path.write_text(markdown, encoding="utf-8")
+    html_path.write_text(markdown_to_html(markdown, report_path, f"{prefix} 基因家族分析结果解读报告"), encoding="utf-8")
+    manifest = {
+        "manifest_version": "1.0",
+        "stage": REPORT_STAGE,
+        "status": "success",
+        "prefix": prefix,
+        "job_id": job_id,
+        "scope_id": scope_id,
+        "analysis_route": analysis_route,
+        "member_set_id": member_set_id,
+        "rnaseq_scope_id": rnaseq_scope_id,
+        "parameters": {"expected_stages": planned, "stage_statuses": {stage: statuses[stage] for stage in planned}},
+        "metrics": {
+            "report_status": report_data.report_status,
+            "generation_mode": generation_mode,
+            "included_stages": list(report_data.included),
+            "unsupported_stages": list(report_data.unsupported_stages),
+            "failed_or_missing_stages": [stage for stage in planned if stage not in report_data.included],
+            "report_renderer_version": REPORT_RENDERER_VERSION,
+            "embedded_figure_count": sum(len(items) for items in figures_by_stage.values()),
+        },
+        "model": model_metadata,
+        "artifacts": {
+            "gene_family_report": {
+                "path": f"final/{report_path.name}",
+                "full_path": str(report_path),
+                "type": "markdown",
+            },
+            "gene_family_report_html": {
+                "path": f"final/{html_path.name}", "full_path": str(html_path), "type": "html",
+            },
+            "gene_family_report_manifest": {
+                "path": f"final/{manifest_path.name}",
+                "full_path": str(manifest_path),
+                "type": "manifest",
+            },
+        },
+    }
+    manifest_path.write_text(json.dumps(manifest, ensure_ascii=False, indent=2) + "\n", encoding="utf-8")
+    return GeneFamilyReportResult(
+        report_path,
+        manifest_path,
+        report_data.report_status,
+        generation_mode,
+        report_data.included,
+        html_path=html_path,
+    )

gfa_core/gene_family_report_data.py

@@ -0,0 +1,767 @@
+"""Deterministic data model for the formal gene-family report."""
+
+from __future__ import annotations
+
+from dataclasses import dataclass
+import csv
+import json
+from pathlib import Path
+import re
+from typing import Any, Mapping, Sequence
+
+
+@dataclass(frozen=True)
+class MissingMetric:
+    """A metric that is genuinely absent from every declared source."""
+
+    label: str
+    checked_sources: tuple[str, ...]
+
+    def render(self) -> str:
+        checked = "、".join(self.checked_sources)
+        return f"{self.label}：源结果未提供此指标（已检查：{checked}）。"
+
+
+@dataclass(frozen=True)
+class Sentence:
+    """One immutable report sentence decided by deterministic code."""
+
+    id: str
+    text: str
+    kind: str
+    evidence_ids: tuple[str, ...] = ()
+
+
+@dataclass(frozen=True)
+class ReportSection:
+    section_id: str
+    stage: str | None
+    facts: tuple[Sentence, ...] = ()
+    observations: tuple[Sentence, ...] = ()
+    interpretations: tuple[Sentence, ...] = ()
+    limitations: tuple[Sentence, ...] = ()
+    recommendations: tuple[Sentence, ...] = ()
+
+    @property
+    def sentences(self) -> tuple[Sentence, ...]:
+        return (
+            self.facts
+            + self.observations
+            + self.interpretations
+            + self.limitations
+            + self.recommendations
+        )
+
+
+@dataclass(frozen=True)
+class ReportData:
+    prefix: str
+    planned: tuple[str, ...]
+    statuses: Mapping[str, str]
+    included: tuple[str, ...]
+    unsupported_stages: tuple[str, ...]
+    report_status: str
+    sections: tuple[ReportSection, ...]
+
+
+STAGE_SECTION_IDS = {
+    "identify": "stage.identify",
+    "phylogeny": "stage.phylogeny",
+    "reference_classification": "stage.reference_classification",
+    "motif": "stage.motif",
+    "domain": "stage.domain",
+    "structure": "stage.structure",
+    "combo": "stage.combo",
+    "promoter": "stage.promoter",
+    "tfbs": "stage.tfbs",
+    "synteny": "stage.synteny",
+    "karyotype": "stage.karyotype",
+    "interspecies_synteny": "stage.interspecies_synteny",
+    "pep": "stage.pep",
+}
+
+
+_REQUIRED_METRICS = {
+    "identify": (("member_count", "最终成员数"),),
+    "phylogeny": (
+        ("sequences_count", "序列数"),
+        ("msa_length", "多序列比对长度"),
+    ),
+    "motif": (
+        ("motif_count", "Motif 类别数"),
+        ("total_sites", "Motif 总位点数"),
+        ("motif_width_range", "Motif 宽度范围"),
+    ),
+    "pep": (
+        ("protein_count", "蛋白数"),
+        ("tasks_success", "成功子任务数"),
+        ("tasks_total", "子任务总数"),
+    ),
+}
+
+
+_ARTIFACT_METRIC_ROLES = {
+    ("identify", "member_count"): ("identified_members", "identified_members_id", "member_ids"),
+    ("pep", "protein_count"): ("pep_properties", "properties_table"),
+}
+
+
+def _manifest_source(manifest: Mapping[str, Any], field: str) -> str:
+    path = str(manifest.get("_manifest_path") or "stage manifest")
+    return f"{Path(path).name}#metrics.{field}"
+
+
+def _artifact_path(final_dir: Path, artifact: Any) -> Path | None:
+    if not isinstance(artifact, Mapping):
+        return None
+    raw = artifact.get("full_path") or artifact.get("path")
+    if not raw:
+        return None
+    path = Path(str(raw))
+    if path.is_file():
+        return path
+    candidates = [final_dir / path.name]
+    if path.is_absolute():
+        job_index = next((index for index, part in enumerate(path.parts) if part.startswith("job_")), None)
+        if job_index is not None:
+            candidates.append(final_dir.parent.joinpath(*path.parts[job_index + 1 :]))
+    else:
+        candidates.append(final_dir.parent / path)
+    return next((candidate for candidate in candidates if candidate.is_file()), None)
+
+
+def _artifact_by_suffix(final_dir: Path, manifest: Mapping[str, Any], suffix: str) -> Path | None:
+    artifacts = manifest.get("artifacts") if isinstance(manifest.get("artifacts"), Mapping) else {}
+    for role in sorted(artifacts):
+        if str(role).endswith(suffix):
+            path = _artifact_path(final_dir, artifacts[role])
+            if path is not None:
+                return path
+    return None
+
+
+def _count_named_records(path: Path) -> int:
+    suffix = path.suffix.lower()
+    try:
+        if suffix in {".fa", ".faa", ".fasta", ".pep"}:
+            return sum(1 for line in path.read_text(encoding="utf-8").splitlines() if line.startswith(">"))
+        if suffix in {".tsv", ".csv"}:
+            delimiter = "\t" if suffix == ".tsv" else ","
+            with path.open("r", encoding="utf-8", newline="") as handle:
+                return sum(1 for _ in csv.DictReader(handle, delimiter=delimiter))
+        if suffix == ".json":
+            value = json.loads(path.read_text(encoding="utf-8"))
+            if isinstance(value, list):
+                return len(value)
+            if isinstance(value, Mapping):
+                for key in ("members", "proteins", "records"):
+                    if isinstance(value.get(key), list):
+                        return len(value[key])
+                raise ValueError("JSON does not contain a declared record list")
+        return sum(1 for line in path.read_text(encoding="utf-8").splitlines() if line.strip())
+    except (OSError, UnicodeError, csv.Error, json.JSONDecodeError, ValueError) as exc:
+        raise ValueError(f"无法解析声明的报告数据源 {path}: {exc}") from exc
+
+
+def _metric(
+    final_dir: Path,
+    stage: str,
+    manifest: Mapping[str, Any],
+    field: str,
+    label: str,
+) -> tuple[Any | MissingMetric, tuple[str, ...]]:
+    metrics = manifest.get("metrics") if isinstance(manifest.get("metrics"), Mapping) else {}
+    manifest_source = _manifest_source(manifest, field)
+    if stage == "identify" and field == "member_count":
+        for alias in ("member_count", "identified_members_count", "final_members"):
+            if metrics.get(alias) is not None:
+                source = _manifest_source(manifest, alias)
+                return metrics[alias], (source,)
+    elif field == "motif_width_range":
+        motifs = metrics.get("motifs")
+        if isinstance(motifs, Sequence) and not isinstance(motifs, (str, bytes)):
+            widths = [item.get("width") for item in motifs if isinstance(item, Mapping) and item.get("width") is not None]
+            if widths:
+                try:
+                    numeric = [int(value) for value in widths]
+                except (TypeError, ValueError) as exc:
+                    raise ValueError(f"无法解析 {manifest_source.replace(field, 'motifs[].width')}") from exc
+                source = _manifest_source(manifest, "motifs[].width")
+                return (min(numeric), max(numeric)), (source,)
+    elif field in metrics and metrics[field] is not None:
+        return metrics[field], (manifest_source,)
+
+    checked = [manifest_source]
+    artifacts = manifest.get("artifacts") if isinstance(manifest.get("artifacts"), Mapping) else {}
+    for role in _ARTIFACT_METRIC_ROLES.get((stage, field), ()):
+        checked.append(f"artifact:{role}")
+        path = _artifact_path(final_dir, artifacts.get(role))
+        if path is not None:
+            return _count_named_records(path), (str(path),)
+    return MissingMetric(label, tuple(checked)), tuple(checked)
+
+
+def _gene_aliases(gene_id: str) -> tuple[str, ...]:
+    value = str(gene_id).strip()
+    aliases = {value, re.sub(r"\.T\d+$", "", value), re.sub(r"\.\d+$", "", value)}
+    return tuple(alias for alias in aliases if alias)
+
+
+def _parse_interspecies_artifacts(
+    final_dir: Path,
+    manifest: Mapping[str, Any],
+) -> tuple[int, int, tuple[str, ...]] | None:
+    artifacts = manifest.get("artifacts") if isinstance(manifest.get("artifacts"), Mapping) else {}
+    mapping_path = _artifact_path(final_dir, artifacts.get("gene_to_species"))
+    collinearity_path = _artifact_by_suffix(final_dir, manifest, "_collinearity")
+    if mapping_path is None or collinearity_path is None:
+        return None
+    groups: dict[str, str] = {}
+    try:
+        with mapping_path.open("r", encoding="utf-8", newline="") as handle:
+            for row in csv.DictReader(handle, delimiter="\t"):
+                gene_id = str(row.get("gene_id") or "").strip()
+                group = str(row.get("group") or "").strip().lower()
+                if gene_id and group in {"target", "ref"}:
+                    for alias in _gene_aliases(gene_id):
+                        groups[alias] = group
+        alignment_blocks = 0
+        target_ref_pairs = 0
+        with collinearity_path.open("r", encoding="utf-8") as handle:
+            for line in handle:
+                if line.startswith("## Alignment"):
+                    alignment_blocks += 1
+                    continue
+                match = re.search(r":\s*(\S+)\s+(\S+)", line) if line.startswith(" ") else None
+                if not match:
+                    continue
+                pair_groups = {
+                    groups[alias]
+                    for gene_id in match.groups()
+                    for alias in _gene_aliases(gene_id)
+                    if alias in groups
+                }
+                if pair_groups == {"target", "ref"}:
+                    target_ref_pairs += 1
+    except (OSError, UnicodeError, csv.Error) as exc:
+        raise ValueError(f"无法解析种间共线性命名产物：{mapping_path}；{collinearity_path}: {exc}") from exc
+    return alignment_blocks, target_ref_pairs, (str(collinearity_path), str(mapping_path))
+
+
+def _sentence(sentence_id: str, text: str, kind: str, *evidence_ids: str) -> Sentence:
+    return Sentence(sentence_id, text, kind, tuple(evidence_ids))
+
+
+def _metrics(manifest: Mapping[str, Any]) -> Mapping[str, Any]:
+    value = manifest.get("metrics")
+    return value if isinstance(value, Mapping) else {}
+
+
+def _artifacts(manifest: Mapping[str, Any]) -> Mapping[str, Any]:
+    value = manifest.get("artifacts")
+    return value if isinstance(value, Mapping) else {}
+
+
+def _artifact_for_roles(
+    final_dir: Path,
+    manifest: Mapping[str, Any],
+    *roles: str,
+) -> tuple[Path | None, str | None]:
+    artifacts = _artifacts(manifest)
+    for role in roles:
+        path = _artifact_path(final_dir, artifacts.get(role))
+        if path is not None:
+            return path, role
+    for role, artifact in artifacts.items():
+        if any(token in str(role) for token in roles):
+            path = _artifact_path(final_dir, artifact)
+            if path is not None:
+                return path, str(role)
+    return None, None
+
+
+def _relative_artifact(final_dir: Path, manifest: Mapping[str, Any], *roles: str) -> tuple[str, str] | None:
+    artifacts = _artifacts(manifest)
+    for role in roles:
+        artifact = artifacts.get(role)
+        path = _artifact_path(final_dir, artifact)
+        if path is not None:
+            try:
+                relative = path.relative_to(final_dir.parent)
+            except ValueError:
+                relative = Path("final") / path.name
+            return relative.as_posix(), f"artifact:{role}"
+        if isinstance(artifact, Mapping):
+            raw = artifact.get("path")
+            if raw and not Path(str(raw)).is_absolute() and ".." not in Path(str(raw)).parts:
+                return Path(str(raw)).as_posix(), f"artifact:{role}"
+    return None
+
+
+def _tabular_rows(path: Path) -> list[dict[str, str]]:
+    try:
+        lines = [
+            line
+            for line in path.read_text(encoding="utf-8").splitlines()
+            if line.strip() and not line.lstrip().startswith("#")
+        ]
+        if not lines:
+            return []
+        return [dict(row) for row in csv.DictReader(lines, delimiter="\t")]
+    except (OSError, UnicodeError, csv.Error) as exc:
+        raise ValueError(f"无法解析命名表格 {path}: {exc}") from exc
+
+
+def _element_counts(path: Path) -> list[tuple[str, int]]:
+    result: list[tuple[str, int]] = []
+    try:
+        for line in path.read_text(encoding="utf-8").splitlines():
+            parts = line.split("\t", 1)
+            if len(parts) != 2:
+                continue
+            try:
+                count = int(parts[0])
+            except ValueError:
+                continue
+            result.append((parts[1].strip(), count))
+    except (OSError, UnicodeError) as exc:
+        raise ValueError(f"无法解析启动子元件计数 {path}: {exc}") from exc
+    return sorted(result, key=lambda item: (-item[1], item[0]))
+
+
+def _key_artifact_sentence(
+    final_dir: Path,
+    manifest: Mapping[str, Any],
+    section_id: str,
+    label: str,
+    *roles: str,
+) -> Sentence | None:
+    artifact = _relative_artifact(final_dir, manifest, *roles)
+    if artifact is None:
+        return None
+    path, evidence = artifact
+    return _sentence(f"{section_id}.artifact.{roles[0]}", f"关键产物：{label} `{path}`。", "fact", evidence)
+
+
+def _stage_section(
+    final_dir: Path,
+    prefix: str,
+    stage: str,
+    status: str,
+    manifest: Mapping[str, Any],
+) -> ReportSection:
+    section_id = STAGE_SECTION_IDS[stage]
+    facts: list[Sentence] = []
+    observations: list[Sentence] = []
+    interpretations: list[Sentence] = []
+    limitations: list[Sentence] = []
+    recommendations: list[Sentence] = []
+
+    if status != "success" or str(manifest.get("status") or "") != "success":
+        facts.append(_sentence(f"{section_id}.status", f"阶段状态：`{status}`。", "fact", _manifest_source(manifest, "status")))
+        limitations.append(
+            _sentence(
+                f"{section_id}.not_completed",
+                "该阶段未成功完成，因此本报告不生成结果推断。",
+                "limitation",
+            )
+        )
+        return ReportSection(section_id, stage, tuple(facts), limitations=tuple(limitations))
+
+    metrics = _metrics(manifest)
+    if stage == "identify":
+        member_count, sources = _metric(final_dir, stage, manifest, "member_count", "最终成员数")
+        family = prefix.split(".", 1)[1] if "." in prefix else prefix
+        facts.append(_sentence(f"{section_id}.object", "分析对象：目标物种完整蛋白集中的候选成员。", "fact"))
+        if isinstance(member_count, MissingMetric):
+            facts.append(_sentence(f"{section_id}.missing.member_count", member_count.render(), "fact", *sources))
+        else:
+            hmm = metrics.get("hmm_members")
+            blast = metrics.get("blast_members")
+            search = f"；HMM 与 BLAST 各支持 {hmm} 和 {blast} 个候选" if hmm is not None and blast is not None else ""
+            observations.append(_sentence(
+                f"{section_id}.observation.members",
+                f"目标物种最终成员数：{member_count}。本阶段形成 {member_count} 个 {family}家族成员{search}。",
+                "observation", *sources,
+            ))
+        interpretations.append(_sentence(f"{section_id}.interpretation", "解释：最终名单是后续阶段的目标物种成员集合；搜索方法的一致数量支持集合稳定，但不等同于实验功能确认。", "interpretation"))
+        limitations.append(_sentence(f"{section_id}.limitation", "限制：成员边界仍受蛋白注释、搜索阈值和转录本选择规则影响。", "limitation"))
+        artifact = _key_artifact_sentence(final_dir, manifest, section_id, "目标成员 FASTA", "identified_members", "target_Ft_identified_members_pep")
+    elif stage == "phylogeny":
+        sequences = metrics.get("sequences_count")
+        msa_length = metrics.get("msa_length")
+        species = metrics.get("scope_species_counts") if isinstance(metrics.get("scope_species_counts"), Mapping) else {}
+        object_text = (
+            f"分析对象：目标物种 {species.get('Ft')} 条与参考物种 {species.get('At')} 条蛋白合并的 {sequences} 条序列。"
+            if species
+            else f"分析对象：目标与参考成员合并集合中的 {sequences} 条蛋白序列。"
+        )
+        facts.append(_sentence(f"{section_id}.object", object_text, "fact", _manifest_source(manifest, "scope_species_counts")))
+        observations.append(_sentence(f"{section_id}.observation", f"序列数：{sequences}；多序列比对长度：{msa_length}。目标与参考成员共 {sequences} 条序列进入比对，本阶段使用 {sequences} 条序列，比对长度为 {msa_length}。", "observation", _manifest_source(manifest, "sequences_count"), _manifest_source(manifest, "msa_length")))
+        interpretations.append(_sentence(f"{section_id}.interpretation", "解释：该树用于比较序列关系并辅助分组；它分析的是目标加参考集合，不应与 24 个目标成员的鉴定集合混称。", "interpretation"))
+        limitations.append(_sentence(f"{section_id}.limitation", "限制：源 Manifest 未汇总各主要分支的支持值，不能据此报告小组稳定性等级。", "limitation"))
+        artifact = _key_artifact_sentence(final_dir, manifest, section_id, "关系树", "tree")
+    elif stage == "reference_classification":
+        target_count = int(metrics.get("target_gene_count") or 0)
+        all_rows = int(metrics.get("subfamily_rows") or 0)
+        assignment_path, assignment_role = _artifact_for_roles(final_dir, manifest, "assignment_table")
+        rows = _tabular_rows(assignment_path) if assignment_path else []
+        counts: dict[str, int] = {}
+        subfamilies: set[str] = set()
+        for row in rows:
+            confidence = str(row.get("confidence") or "").lower()
+            if confidence:
+                counts[confidence] = counts.get(confidence, 0) + 1
+            label = str(row.get("assigned_subfamily") or "").strip()
+            if label:
+                subfamilies.add(label)
+        facts.append(_sentence(f"{section_id}.object", f"分析对象：{target_count} 个目标成员与 {max(0, all_rows - target_count)} 个参考成员的联合分类。", "fact"))
+        if rows:
+            observations.append(_sentence(f"{section_id}.observation", f"分类结果：{len(rows)} 个目标成员分入 {len(subfamilies)} 个小组；高信心 {counts.get('high', 0)} 个、中信心 {counts.get('medium', 0)} 个、低信心 {counts.get('low', 0)} 个。", "observation", f"artifact:{assignment_role}"))
+        else:
+            missing = MissingMetric("分类信心分布", ("artifact:assignment_table",))
+            facts.append(_sentence(f"{section_id}.missing", missing.render(), "fact", *missing.checked_sources))
+        interpretations.append(_sentence(f"{section_id}.interpretation", "解释：参考成员为目标成员提供小组标签，其中低信心标签只能作为后续分组线索。", "interpretation"))
+        limitations.append(_sentence(f"{section_id}.limitation", "限制：分类依赖参考成员覆盖和树拓扑；低信心成员不宜用于强分组结论。", "limitation"))
+        artifact = _key_artifact_sentence(final_dir, manifest, section_id, "分类明细", "assignment_table")
+    elif stage == "motif":
+        motifs = metrics.get("motifs") if isinstance(metrics.get("motifs"), Sequence) else []
+        widths = [int(item["width"]) for item in motifs if isinstance(item, Mapping) and item.get("width") is not None]
+        count = metrics.get("motif_count")
+        total_sites = metrics.get("total_sites")
+        sequence_count = metrics.get("sequence_count") or metrics.get("protein_record_count")
+        if sequence_count is None:
+            facts.append(_sentence(f"{section_id}.object", "分析对象：当前 Motif Manifest 声明的蛋白集合。", "fact"))
+            missing = MissingMetric("Motif 输入序列数", (_manifest_source(manifest, "sequence_count"),))
+            facts.append(_sentence(f"{section_id}.missing.sequence_count", missing.render(), "fact", *missing.checked_sources))
+        else:
+            facts.append(_sentence(f"{section_id}.object", f"分析对象：目标与参考成员合并集合中的 {sequence_count} 条蛋白序列。", "fact"))
+        if widths:
+            observations.append(_sentence(f"{section_id}.observation", f"共识别 {count} 个 motif，总位点数为 {total_sites}，宽度范围为 {min(widths)}–{max(widths)}。", "observation", _manifest_source(manifest, "motifs")))
+        else:
+            missing = MissingMetric("Motif 宽度范围", (_manifest_source(manifest, "motifs[].width"),))
+            facts.append(_sentence(f"{section_id}.missing", missing.render(), "fact", *missing.checked_sources))
+        interpretations.append(_sentence(f"{section_id}.interpretation", "解释：这些统计描述短片段的类别、总命中与长度范围；总位点数不能替代逐成员覆盖率。", "interpretation"))
+        limitations.append(_sentence(f"{section_id}.limitation", "限制：源结果未汇总同组排列一致性或异常成员，不能从总数推断特定小组功能分化。", "limitation"))
+        artifact = _key_artifact_sentence(final_dir, manifest, section_id, "Motif 排列图", "motif_plot_png")
+    elif stage == "domain":
+        total = metrics.get("proteins_total")
+        hit = metrics.get("proteins_with_domains")
+        top_domains = metrics.get("top_domains") if isinstance(metrics.get("top_domains"), Sequence) else []
+        top_text = "、".join(f"{item[0]} {item[1]} 次" for item in top_domains if isinstance(item, Sequence) and len(item) >= 2)
+        facts.append(_sentence(f"{section_id}.object", f"分析对象：目标与参考成员合并集合中的 {total} 条蛋白。", "fact"))
+        observations.append(_sentence(f"{section_id}.observation", f"Domain 覆盖为 {hit}/{total} 条蛋白；主要命中为 {top_text}。", "observation", _manifest_source(manifest, "proteins_with_domains"), _manifest_source(manifest, "top_domains")))
+        interpretations.append(_sentence(f"{section_id}.interpretation", "解释：全部蛋白均有 Domain 命中，SBP 的普遍命中支持该合并集合保留家族典型区域；Ig_SPL7 只出现在部分命中记录中。", "interpretation"))
+        limitations.append(_sentence(f"{section_id}.limitation", "限制：命中次数不等于独立蛋白数，Ig_SPL7 的 12 次命中不能直接解释为 12 个功能不同成员。", "limitation"))
+        artifact = _key_artifact_sentence(final_dir, manifest, section_id, "Domain 图", "domain_plot_png")
+    elif stage == "structure":
+        total = metrics.get("genes_total")
+        covered = metrics.get("genes_with_structure")
+        outliers = metrics.get("long_gene_outliers") if isinstance(metrics.get("long_gene_outliers"), Sequence) else []
+        outlier = next((item for item in outliers if isinstance(item, Mapping)), None)
+        facts.append(_sentence(f"{section_id}.object", f"分析对象：目标与参考合并集合的 {total} 个成员；结构注释实际匹配 {covered} 个目标成员。", "fact"))
+        detail = f"；长基因异常 1 个，为 {outlier.get('gene_id')}（{outlier.get('length_bp')} bp，{outlier.get('cds_count')} 个 CDS）" if outlier else ""
+        observations.append(_sentence(f"{section_id}.observation", f"结构覆盖为 {covered}/{total}{detail}。", "observation", _manifest_source(manifest, "genes_with_structure"), _manifest_source(manifest, "long_gene_outliers")))
+        interpretations.append(_sentence(f"{section_id}.interpretation", "解释：24/40 的覆盖差异主要反映参考成员未纳入当前目标物种 GFF3；超长基因需要回到原始注释核对。", "interpretation"))
+        limitations.append(_sentence(f"{section_id}.limitation", "限制：未匹配的参考成员不能作为目标物种结构缺失；单个长基因也不能单独证明功能变化。", "limitation"))
+        artifact = _key_artifact_sentence(final_dir, manifest, section_id, "结构明细", "structure_tsv")
+    elif stage == "combo":
+        parameters = manifest.get("parameters") if isinstance(manifest.get("parameters"), Mapping) else {}
+        panels = [str(value).capitalize() for value in parameters.get("panels", [])]
+        panel_text = "、".join(panels) if panels else "Motif、Domain、Structure"
+        facts.append(_sentence(f"{section_id}.object", "分析对象：目标与参考成员合并集合的关系树及其多类结构证据。", "fact"))
+        observations.append(_sentence(f"{section_id}.observation", f"组合图已将关系树与 {panel_text} 三类面板汇总到同一图中。", "observation", _manifest_source(manifest, "parameters.panels")))
+        facts.append(_sentence(f"{section_id}.missing", "组合证据一致性数值：源结果未提供此指标（已检查：Combo Manifest、combo_plot_png、combo_plot_pdf）。", "fact"))
+        interpretations.append(_sentence(f"{section_id}.interpretation", "解释：该图适合逐行核对多类证据，但当前命名源没有可机器追溯的一致性评分。", "interpretation"))
+        limitations.append(_sentence(f"{section_id}.limitation", "限制：没有逐成员一致性表时，不指定稳定小组、异常成员或候选基因。", "limitation"))
+        artifact = _key_artifact_sentence(final_dir, manifest, section_id, "组合图", "combo_plot_png")
+    elif stage == "promoter":
+        genes = metrics.get("genes_with_elements")
+        total = metrics.get("total_elements")
+        unique = metrics.get("unique_elements")
+        counts_path, counts_role = _artifact_for_roles(final_dir, manifest, "element_counts")
+        counts = _element_counts(counts_path) if counts_path else []
+        category_path, category_role = _artifact_for_roles(final_dir, manifest, "position_distribution_summary", "promoter_category_summary")
+        categories = _tabular_rows(category_path) if category_path else []
+        enrichment_path, enrichment_role = _artifact_for_roles(final_dir, manifest, "enrichment_analysis")
+        enrichment_rows = _tabular_rows(enrichment_path) if enrichment_path else []
+        co_path, co_role = _artifact_for_roles(final_dir, manifest, "cooccurrence_analysis")
+        co_rows = _tabular_rows(co_path) if co_path else []
+        complexity_path, complexity_role = _artifact_for_roles(final_dir, manifest, "complexity_analysis")
+        complexity_rows = _tabular_rows(complexity_path) if complexity_path else []
+        facts.append(_sentence(f"{section_id}.object", f"分析对象：检测到顺式元件的 {genes} 个目标基因启动子。", "fact"))
+        top_elements = "、".join(f"{name} {count} 个" for name, count in counts[:3]) or "源结果未提供主要元件排行"
+        top_category = categories[0] if categories else {}
+        category_name = top_category.get("big7") or top_category.get("Category")
+        category_count = top_category.get("element_instances") or top_category.get("Count")
+        category_text = ""
+        if category_name and category_count and total:
+            category_text = f"；主要类别为 {category_name} {category_count} 个（{int(category_count) / int(total) * 100:.1f}%）"
+        enrichment_top = enrichment_rows[0] if enrichment_rows else {}
+        enrichment_text = ""
+        if enrichment_top:
+            enrichment_text = f"；最高列示富集比为小组 {enrichment_top.get('Subfamily')} 的 {enrichment_top.get('Element')}（{enrichment_top.get('EnrichmentRatio')}）"
+        co_top = co_rows[0] if co_rows else {}
+        co_text = ""
+        if co_top:
+            co_text = f"；最常见共现为 {co_top.get('Element1')}—{co_top.get('Element2')}（{co_top.get('Co-occurrenceCount')} 个基因）"
+        complex_top = complexity_rows[0] if complexity_rows else {}
+        complex_text = ""
+        if complex_top:
+            complex_text = f"；复杂度最高的候选为 {complex_top.get('Gene')}（得分 {complex_top.get('ComplexityScore')}）"
+        observations.append(_sentence(f"{section_id}.observation", f"{genes} 个目标基因共记录 {total} 个元件、{unique} 种类型；主要元件为 {top_elements}{category_text}{enrichment_text}{co_text}{complex_text}。", "observation", *(f"artifact:{role}" for role in (counts_role, category_role, enrichment_role, co_role, complexity_role) if role)))
+        interpretations.append(_sentence(f"{section_id}.interpretation", "解释：类别比例、共现和复杂度描述的是上游序列预测模式；富集比可用于定位值得复核的小组—元件组合。", "interpretation"))
+        limitations.append(_sentence(f"{section_id}.limitation", "限制：仅 23 个目标基因有元件记录，且序列片段出现、共现或高复杂度都不能证明真实调控。", "limitation"))
+        artifact = _key_artifact_sentence(final_dir, manifest, section_id, "启动子汇总图", "viz_Ft.SPL_promoter_v6_combined", "promoter_v6_combined")
+    elif stage == "tfbs":
+        promoters = metrics.get("family_promoter_count")
+        sites = metrics.get("family_site_count")
+        motif_count = metrics.get("family_motif_count")
+        background = metrics.get("background_promoter_count")
+        summary_path, summary_role = _artifact_for_roles(final_dir, manifest, "motif_summary")
+        rows = _tabular_rows(summary_path) if summary_path else []
+        rows.sort(key=lambda row: int(float(row.get("family_sites") or 0)), reverse=True)
+        top = rows[0] if rows else {}
+        facts.append(_sentence(f"{section_id}.object", f"分析对象：{promoters} 条目标成员启动子；Motif 模型来自 {metrics.get('motif_database', '声明数据库')}。", "fact"))
+        top_text = f"；最高频 Motif {top.get('motif_id')} 有 {top.get('family_sites')} 个位点，覆盖 {top.get('family_genes_with_site')} 个基因" if top else ""
+        observations.append(_sentence(f"{section_id}.observation", f"{promoters} 条目标启动子共得到 {sites} 个预测位点、{motif_count} 个有命中的 Motif{top_text}。", "observation", _manifest_source(manifest, "family_site_count"), *(tuple([f"artifact:{summary_role}"]) if summary_role else ())))
+        if int(background or 0) == 0:
+            facts.append(_sentence(f"{section_id}.no_background", "背景启动子为 0，不能据此进行家族相对背景的富集判断。", "fact", _manifest_source(manifest, "background_promoter_count")))
+        interpretations.append(_sentence(f"{section_id}.interpretation", "解释：高频位点表示序列扫描中出现较多，只能作为潜在结合线索，不能写成转录因子已调控目标基因。", "interpretation"))
+        limitations.append(_sentence(f"{section_id}.limitation", "限制：Motif 使用最近分类物种投射且缺少背景集合，统计富集和物种特异调控均无法确认。", "limitation"))
+        artifact = _key_artifact_sentence(final_dir, manifest, section_id, "Motif 汇总表", "motif_summary")
+    elif stage == "synteny":
+        facts.append(_sentence(f"{section_id}.object", f"分析对象：目标物种全基因组 {metrics.get('gene_count')} 个基因，而非仅家族成员。", "fact"))
+        observations.append(_sentence(f"{section_id}.observation", f"全基因组结果记录 {metrics.get('synteny_blocks')} 个共线性区块、{metrics.get('gene_pairs')} 个基因对，并绘制 {metrics.get('circos_chromosomes_plotted')} 条染色体。", "observation", _manifest_source(manifest, "synteny_blocks"), _manifest_source(manifest, "gene_pairs")))
+        facts.append(_sentence(f"{section_id}.missing", "源结果未提供家族专属复制基因对（已检查：Synteny Manifest、collinearity artifact、dup_class_log artifact）。", "fact"))
+        interpretations.append(_sentence(f"{section_id}.interpretation", "解释：这些数字刻画全基因组共线性背景，不能直接归因于 SPL 家族扩张。", "interpretation"))
+        limitations.append(_sentence(f"{section_id}.limitation", "限制：缺少家族成员专属 links 或复制类型表，本节不判断大片段复制或串联复制孰为家族主要来源。", "limitation"))
+        artifact = _key_artifact_sentence(final_dir, manifest, section_id, "Circos 图", "circos_png")
+    elif stage == "karyotype":
+        species = metrics.get("species") if isinstance(metrics.get("species"), Mapping) else {}
+        item = next(iter(species.values()), {}) if species else metrics
+        genes = item.get("gene_count") if isinstance(item, Mapping) else None
+        chromosomes = item.get("chromosome_count") if isinstance(item, Mapping) else None
+        facts.append(_sentence(f"{section_id}.object", f"分析对象：目标物种的 {genes} 个家族成员。", "fact"))
+        observations.append(_sentence(f"{section_id}.observation", f"核型分析将{genes}个家族成员定位到{chromosomes}条染色体。", "observation", _manifest_source(manifest, "species.*.gene_count"), _manifest_source(manifest, "species.*.chromosome_count")))
+        interpretations.append(_sentence(f"{section_id}.interpretation", "解释：染色体位置图确认了目标成员的基因组分布范围，可与家族专属复制证据联合阅读。", "interpretation"))
+        limitations.append(_sentence(f"{section_id}.limitation", "限制：Manifest 未汇总每条染色体的成员数，不能仅凭 24/8 判断分布均匀性或成簇关系。", "limitation"))
+        artifact = _key_artifact_sentence(final_dir, manifest, section_id, "核型图", "Ft_karyotype_png")
+    elif stage == "interspecies_synteny":
+        parsed = _parse_interspecies_artifacts(final_dir, manifest)
+        target = metrics.get("target_family_count")
+        reference = metrics.get("ref_family_count")
+        facts.append(_sentence(f"{section_id}.object", f"分析对象：目标物种 {target} 个与参考物种 {reference} 个家族成员，并结合两物种全基因组共线性。", "fact"))
+        if parsed is None:
+            missing = MissingMetric(
+                "种间共线性区块与家族成员配对",
+                ("artifact:gene_to_species", "artifact:*_collinearity"),
+            )
+            facts.append(
+                _sentence(
+                    f"{section_id}.missing.collinearity_summary",
+                    missing.render(),
+                    "fact",
+                    *missing.checked_sources,
+                )
+            )
+        else:
+            alignment_blocks, target_ref_pairs, sources = parsed
+            observations.append(
+                _sentence(
+                    f"{section_id}.observation.summary",
+                    f"共线性区块数：{alignment_blocks}（全基因组 block {alignment_blocks} 个）；"
+                    f"家族成员目标—参考配对数：{target_ref_pairs}（目标-参考家族 {target_ref_pairs} 对）。",
+                    "observation",
+                    *sources,
+                )
+            )
+        interpretations.append(_sentence(f"{section_id}.interpretation", "解释：全基因组区块提供跨物种背景，2 个目标—参考家族配对是当前命名产物中可直接追溯的家族关系。", "interpretation"))
+        limitations.append(_sentence(f"{section_id}.limitation", "限制：共线性配对只能支持保留关系线索，不能直接移植参考物种功能或断言一一对应。", "limitation"))
+        artifact = _key_artifact_sentence(final_dir, manifest, section_id, "标记版种间共线性图", "synteny_marked_png")
+    elif stage == "pep":
+        protein_count = metrics.get("protein_count")
+        success = metrics.get("tasks_success")
+        total_tasks = metrics.get("tasks_total")
+        counts = metrics.get("subcellular_counts") if isinstance(metrics.get("subcellular_counts"), Mapping) else {}
+        facts.append(_sentence(f"{section_id}.object", f"分析对象：目标与参考成员合并集合中的 {protein_count} 条蛋白。", "fact"))
+        localization = ""
+        if counts:
+            localization = f"；亚细胞定位预测为 Nucleus {counts.get('Nucleus', 0)}、Cytoplasm {counts.get('Cytoplasm', 0)}、Chloroplast {counts.get('Chloroplast', 0)}"
+        else:
+            missing = MissingMetric("亚细胞定位分类计数", (_manifest_source(manifest, "subcellular_counts"),))
+            facts.append(_sentence(f"{section_id}.missing.localization", missing.render(), "fact", *missing.checked_sources))
+        observations.append(_sentence(f"{section_id}.observation", f"蛋白数：{protein_count}；子任务成功数：{success}/{total_tasks}。本阶段覆盖 {protein_count} 个蛋白，子任务成功数为 {success}/{total_tasks}{localization}。", "observation", _manifest_source(manifest, "protein_count"), _manifest_source(manifest, "subcellular_counts")))
+        interpretations.append(_sentence(f"{section_id}.interpretation", "解释：核定位预测占多数，与转录因子研究方向相容，但这里只能作为定位线索。", "interpretation"))
+        limitations.append(_sentence(f"{section_id}.limitation", "限制：理化性质、跨膜与亚细胞定位均为计算预测；当前汇总不能替代逐蛋白实验验证。", "limitation"))
+        artifact = _key_artifact_sentence(final_dir, manifest, section_id, "亚细胞定位表", "subcellular_data")
+    else:
+        artifact = None
+
+    if artifact is not None:
+        facts.append(artifact)
+    return ReportSection(
+        section_id,
+        stage,
+        tuple(facts),
+        tuple(observations),
+        tuple(interpretations),
+        tuple(limitations),
+        tuple(recommendations),
+    )
+
+
+def build_report_data(
+    final_dir: Path | str,
+    prefix: str,
+    planned: Sequence[str],
+    statuses: Mapping[str, str],
+    manifests: Mapping[str, Mapping[str, Any]],
+) -> ReportData:
+    """Build the only deterministic data source used by both render modes."""
+
+    final_path = Path(final_dir)
+    unsupported_stages = tuple(stage for stage in planned if stage not in STAGE_SECTION_IDS)
+    included = tuple(
+        stage
+        for stage in planned
+        if stage in STAGE_SECTION_IDS
+        and statuses.get(stage) == "success"
+        and str((manifests.get(stage) or {}).get("status") or "") == "success"
+    )
+    report_status = "partial" if len(included) != len(planned) else "final"
+    success_count = len(included)
+    failed_count = len(planned) - success_count
+    stage_sections = [
+        _stage_section(final_path, prefix, stage, str(statuses.get(stage) or "missing"), manifests.get(stage) or {})
+        for stage in planned
+        if stage in STAGE_SECTION_IDS
+    ]
+    stage_sections_by_name = {section.stage: section for section in stage_sections}
+    summary_texts = [
+        f"本次按用户计划检查 {len(planned)} 个阶段，其中 {success_count} 个成功结果纳入正文。",
+        f"报告状态为{'最终报告' if report_status == 'final' else '部分报告'}，未纳入阶段数为 {failed_count}。",
+    ]
+    for stage in (
+        "identify", "phylogeny", "reference_classification", "domain", "structure",
+        "promoter", "karyotype", "pep",
+    ):
+        section = stage_sections_by_name.get(stage)
+        if section and section.observations and len(summary_texts) < 9:
+            summary_texts.append(section.observations[0].text.replace("。", "；").rstrip("；") + "。")
+    for stage in ("identify", "phylogeny"):
+        section = stage_sections_by_name.get(stage)
+        if section:
+            for sentence in section.facts:
+                if sentence.text.startswith("分析对象：") and len(summary_texts) < 5:
+                    summary_texts.append(sentence.text)
+    while len(summary_texts) < 5:
+        summary_texts.append("正文只纳入计划内成功阶段，并在各节分别声明实际分析对象。")
+    summary_texts.append("本摘要只组合下文可追溯的阶段事实，不据此指定候选基因或推断功能因果。")
+    summary_facts = tuple(
+        _sentence(f"common.front.summary.{index}", text, "fact")
+        for index, text in enumerate(summary_texts[:10], start=1)
+    )
+    completion_facts = [
+        _sentence(
+            "common.completeness.counts",
+            f"本次计划 {len(planned)} 个阶段，纳入解读 {success_count} 个，未纳入 {failed_count} 个。",
+            "fact",
+        )
+    ]
+    for stage in planned:
+        supported = stage in STAGE_SECTION_IDS
+        completion_facts.append(
+            _sentence(
+                f"common.completeness.stage.{stage}",
+                f"计划阶段 `{stage}`：源状态 `{statuses.get(stage, 'missing')}`；"
+                f"模板支持：{'是' if supported else '不支持'}；"
+                f"纳入解读：{'是' if stage in included else '否'}。",
+                "fact",
+            )
+        )
+    sections: list[ReportSection] = [
+        ReportSection(
+            "common.front",
+            None,
+            facts=summary_facts,
+        ),
+        ReportSection(
+            "common.route",
+            None,
+            facts=(
+                _sentence("common.route.stages", f"本报告按计划阶段顺序解读：{'、'.join(planned)}。", "fact"),
+            ),
+        ),
+        ReportSection(
+            "common.completeness",
+            None,
+            facts=tuple(completion_facts),
+        ),
+    ]
+    sections.extend(stage_sections)
+    integrated_observations = [
+        section.observations[0]
+        for section in stage_sections
+        if section.observations
+    ][:3]
+    sections.extend(
+        [
+            ReportSection(
+                "common.integrated",
+                None,
+                facts=tuple(
+                    _sentence(
+                        f"common.integrated.fact.{index}",
+                        sentence.text,
+                        "fact",
+                        *sentence.evidence_ids,
+                    )
+                    for index, sentence in enumerate(integrated_observations, start=1)
+                ),
+                interpretations=(),
+            ),
+            ReportSection(
+                "common.limitations",
+                None,
+                limitations=(
+                    _sentence(
+                        "common.limitations.general",
+                        "基因组与注释质量、分析阈值和计算预测边界都会影响结果，缺少实验信息时不能确定具体生物学功能。",
+                        "limitation",
+                    ),
+                ),
+            ),
+            ReportSection(
+                "common.recommendations",
+                None,
+                recommendations=(
+                    _sentence(
+                        "common.recommendations.general",
+                        "优先核对异常成员的序列与注释，并用表达、定位或功能实验验证由至少两类独立证据支持的候选。",
+                        "recommendation",
+                    ),
+                ),
+            ),
+            ReportSection(
+                "common.checklist",
+                None,
+                recommendations=(
+                    _sentence(
+                        "common.checklist.traceability",
+                        "交付前应确认每项结论均能追溯到当前身份的 Manifest 或命名结果文件。",
+                        "recommendation",
+                    ),
+                ),
+            ),
+        ]
+    )
+    return ReportData(
+        prefix,
+        tuple(planned),
+        dict(statuses),
+        included,
+        unsupported_stages,
+        report_status,
+        tuple(sections),
+    )

gfa_core/gene_family_report_html.py

@@ -0,0 +1,122 @@
+# HTML rendering and representative-figure selection for gene-family reports.
+from __future__ import annotations
+import base64
+import html
+import mimetypes
+from pathlib import Path
+import re
+from typing import Any, Iterable, Mapping
+
+REPORT_RENDERER_VERSION = "gene_family_report_html_v4"
+IMAGE_SUFFIXES = {".png", ".jpg", ".jpeg", ".webp", ".svg"}
+STAGE_FIGURE_TEXT = {
+    "phylogeny": ("系统发育图展示目标成员与参考成员之间的聚类关系。", "结合分支拓扑、节点支持率和成员标签阅读，不把邻近关系直接等同于功能一致。", "目标与参考序列的聚类关系见图，具体分支以图中节点和标签为准。"),
+    "reference_classification": ("参考分类图展示目标成员与已知参考亚家族的相对位置。", "结合参考成员标签、目标成员所在分支及支持率判断分类线索。", "目标成员的参考辅助分类位置见图，分类结论以本节确定性结果为准。"),
+    "motif": ("Motif图展示各成员短保守片段的组成、顺序和相对位置。", "按成员逐行比较Motif颜色、编号、数量和排列，不把Motif相似直接写成功能相同。", "本次成员的Motif组成与排列见图，数量统计以本节结果数据为准。"),
+    "domain": ("Domain图展示蛋白主要功能区域的类型和位置。", "结合结构域名称、覆盖区间和成员间差异阅读。", "各成员的结构域组成见图，覆盖数量和比例以本节结果数据为准。"),
+    "structure": ("基因结构图展示外显子、内含子和基因长度的相对关系。", "结合图例比较外显子数量、结构排列和异常长基因。", "目标成员的基因结构差异见图，具体统计以本节结果数据为准。"),
+    "combo": ("综合组合图并列展示系统发育、Motif、Domain和基因结构信息。", "沿同一成员横向比较多个证据面板，关注一致模式与明显差异。", "多类结构证据的对应关系见图，不据此单独确认功能改变。"),
+    "promoter": ("启动子图展示顺式作用元件的类型、数量或位置分布。", "结合元件图例、位置坐标和成员标签比较，不把预测位点写成已验证调控。", "目标成员启动子元件的分布见图，数量和类别统计以本节结果数据为准。"),
+    "synteny": ("共线性图展示染色体区段和基因对之间的连线关系。", "结合染色体标签、连线和家族成员标记阅读，区分全基因组背景与家族证据。", "本次共线性背景见图，家族专属关系以本节可追溯结果为准。"),
+    "karyotype": ("染色体分布图展示家族成员在染色体上的物理位置。", "结合染色体编号、位置刻度和基因标签查看分布范围。", "目标家族成员的染色体位置见图，成员数和染色体数以本节结果数据为准。"),
+    "interspecies_synteny": ("种间共线性图展示目标物种与参考物种之间的保守区块和基因连线。", "结合物种、染色体、标记连线和成员标签阅读，不直接移植参考物种功能。", "跨物种共线性关系见图，目标与参考家族配对数以本节结果数据为准。"),
+    "pep": ("蛋白特征图展示理化性质、跨膜区域或亚细胞定位预测结果。", "根据具体图例比较成员分布，区分不同预测子任务。", "本次蛋白预测结果见图，具体数量与分类以本节结果数据为准。"),
+    "tfbs": ("Motif统计图展示不同基因的位点计数、全部Motif计数和高频Motif。", "结合基因标签、颜色图例、计数刻度和Motif名称比较分布。", "本次gene、all和top motif统计结果见图，具体计数以图例和源结果为准。"),
+}
+
+def _artifact_image(final_dir: Path, artifact: Any) -> Path | None:
+    if not isinstance(artifact, Mapping): return None
+    raw = str(artifact.get("full_path") or artifact.get("path") or "").strip()
+    target = final_dir / Path(raw).name if raw else None
+    return target if target and target.is_file() and target.suffix.lower() in IMAGE_SUFFIXES else None
+
+def _kind(role: str, filename: str) -> str:
+    value = f"{role} {filename}".lower()
+    groups = (("gene_motif_count", ("gene_motif_heatmap",)), ("all_motif_count", ("all_motif_count_heatmap",)), ("top_motif", ("top_motifs_barplot",)), ("tree", ("tree",)), ("msa", ("msa_plot", "conserved_region")), ("synteny", ("synteny", "circos")), ("motif", ("motif",)), ("domain", ("domain",)), ("structure", ("structure",)), ("combo", ("combo",)), ("promoter_combined", ("promoter_v6_combined",)), ("promoter_position", ("promoter_position_distribution",)), ("promoter", ("promoter",)), ("karyotype", ("karyotype",)), ("tmhmm", ("tmhmm",)), ("subcellular", ("subcellular",)), ("properties", ("properties",)))
+    return next((kind for kind, tokens in groups if any(token in value for token in tokens)), role.lower())
+
+def _rank(role: str, path: Path) -> tuple[int, int, str]:
+    value = f"{role} {path.name}".lower()
+    role_lower = role.lower()
+    preferred = -5 if role_lower == "synteny_marked_png" else -4 if "circular" in value else -2 if "marked" in value else -1 if "summary" in value else 0
+    return (preferred, 0, path.name)
+
+def collect_stage_figures(final_dir: Path, stages: Iterable[str], manifests: Mapping[str, Mapping[str, Any]]) -> dict[str, list[dict[str, str]]]:
+    result = {}
+    for stage in stages:
+        selected = {}; seen = set()
+        for role, artifact in ((manifests.get(stage) or {}).get("artifacts") or {}).items():
+            target = _artifact_image(final_dir, artifact)
+            if target is None or target.name in seen: continue
+            seen.add(target.name); kind = _kind(str(role), target.name)
+            candidate = (_rank(str(role), target), {"role": str(role), "filename": target.name})
+            if kind not in selected or candidate[0] < selected[kind][0]: selected[kind] = candidate
+        if stage == "tfbs":
+            for role, pattern in (
+                ("gene_motif_heatmap", "*gene_motif_heatmap.png"),
+                ("all_motif_count_heatmap", "*all_motif_count_heatmap.png"),
+                ("top_motifs_barplot", "*top_motifs_barplot.png"),
+            ):
+                matches = sorted(final_dir.glob(pattern))
+                if not matches:
+                    continue
+                target = matches[0]
+                kind = _kind(role, target.name)
+                selected[kind] = (_rank(role, target), {"role": role, "filename": target.name})
+        figures = [entry[1] for _, entry in sorted(selected.items(), key=lambda item: (item[1][0], item[0]))][:4]
+        if figures: result[stage] = figures
+    return result
+
+def render_stage_figures(stage: str, figures: list[dict[str, str]]) -> str:
+    meaning, reading, result = STAGE_FIGURE_TEXT.get(stage, ("结果图展示该阶段的主要计算产物。", "结合图题、坐标轴、图例和成员标签阅读。", "本阶段代表结果见图，具体事实以本节确定性数据为准。"))
+    lines = ["", "### 结果图"]
+    for index, figure in enumerate(figures, 1):
+        title = figure["role"]
+        lines.extend(["", f"#### 图 {index}. {title}", "", f"![{title}]({figure['filename']})", "", f"图形含义：{meaning}", "", f"阅读方法：{reading}", "", f"本次结果：{result}"])
+    return "\n".join(lines)
+
+def _inline(value: str, report_path: Path) -> str:
+    tokens = []
+    def stash(content):
+        token = f"@@GFAHTML{len(tokens)}@@"; tokens.append(content); return token
+    def image(match):
+        alt = match.group(1).strip() or "基因家族结果图"; target = report_path.parent / Path(match.group(2).strip().replace("\\", "/")).name
+        source = target.name; mime = mimetypes.guess_type(target.name)[0] or ""
+        if target.is_file() and mime.startswith("image/"): source = f"data:{mime};base64,{base64.b64encode(target.read_bytes()).decode('ascii')}"
+        return stash(f'<figure><img src="{html.escape(source, quote=True)}" alt="{html.escape(alt, quote=True)}" loading="lazy"><figcaption>{html.escape(alt)}</figcaption></figure>')
+    def link(match):
+        label = html.escape(match.group(1).strip()); target = Path(match.group(2).strip().replace("\\", "/")).name
+        return stash(f'<a href="{html.escape(target, quote=True)}" target="_blank" rel="noopener">{label}</a>')
+    value = re.sub(r"!\[([^\]]*)\]\(([^)]+)\)", image, value); value = re.sub(r"(?<!!)\[([^\]]+)\]\(([^)]+)\)", link, value)
+    value = re.sub(r"`([^`]+)`", lambda m: stash(f"<code>{html.escape(m.group(1))}</code>"), value); value = html.escape(value); value = re.sub(r"\*\*([^*]+)\*\*", r"<strong>\1</strong>", value)
+    for index, content in enumerate(tokens): value = value.replace(f"@@GFAHTML{index}@@", content)
+    return value
+
+def markdown_to_html(markdown: str, report_path: Path, title: str) -> str:
+    output = []; list_tag = ""; table_rows = []
+    def close_list():
+        nonlocal list_tag
+        if list_tag: output.append(f"</{list_tag}>"); list_tag = ""
+    def flush_table():
+        nonlocal table_rows
+        if not table_rows: return
+        rows, table_rows = table_rows, []
+        if len(rows) >= 2 and all(re.fullmatch(r":?-{3,}:?", cell.replace(" ", "")) for cell in rows[1]):
+            output.append('<div class="table-wrap"><table><thead><tr>' + ''.join(f"<th>{_inline(c, report_path)}</th>" for c in rows[0]) + '</tr></thead><tbody>'); output.extend('<tr>' + ''.join(f"<td>{_inline(c, report_path)}</td>" for c in row) + '</tr>' for row in rows[2:]); output.append('</tbody></table></div>')
+        else: output.extend(f"<p>{_inline(' | '.join(row), report_path)}</p>" for row in rows)
+    for raw in markdown.splitlines():
+        line = raw.strip()
+        if line.startswith("|") and line.endswith("|"): close_list(); table_rows.append([c.strip() for c in line.strip("|").split("|")]); continue
+        flush_table()
+        if not line: close_list(); continue
+        heading = re.match(r"^(#{1,6})\s+(.+)$", line)
+        if heading: close_list(); level=len(heading.group(1)); output.append(f"<h{level}>{_inline(heading.group(2), report_path)}</h{level}>"); continue
+        if line.startswith("> "): close_list(); output.append(f"<blockquote>{_inline(line[2:], report_path)}</blockquote>"); continue
+        item = re.match(r"^([-*]|\d+\.)\s+(.+)$", line)
+        if item:
+            tag = "ol" if item.group(1)[0].isdigit() else "ul"
+            if list_tag != tag: close_list(); list_tag=tag; output.append(f"<{tag}>")
+            output.append(f"<li>{_inline(item.group(2), report_path)}</li>"); continue
+        close_list(); rendered=_inline(line, report_path); output.append(rendered if line.startswith("![") else f"<p>{rendered}</p>")
+    close_list(); flush_table(); body="\n".join(output); safe_title=html.escape(title)
+    return f'''<!doctype html><html lang="zh-CN"><head><meta charset="utf-8"><meta name="viewport" content="width=device-width,initial-scale=1"><title>{safe_title}</title><style>:root{{--ink:#17352c;--paper:#fffdf7;--line:#dce5dc;--accent:#157a57;--soft:#edf6ef;--warm:#f2e5c7}}*{{box-sizing:border-box}}html{{background:#e9efe9}}body{{margin:0;color:var(--ink);font-family:"Noto Serif SC","Source Han Serif SC","Microsoft YaHei",serif;line-height:1.85;background:radial-gradient(circle at top left,#f8edcf 0,transparent 28rem),linear-gradient(150deg,#edf4ec,#f6f2e8)}}main{{width:min(1120px,calc(100% - 32px));margin:32px auto 64px;padding:clamp(28px,5vw,72px);background:var(--paper);border:1px solid rgba(23,53,44,.12);border-radius:22px;box-shadow:0 24px 70px rgba(35,62,48,.14);overflow:hidden}}h1{{font-size:clamp(2rem,5vw,3.5rem);line-height:1.2;color:#0d4f39}}h2{{margin:3.2rem 0 1.1rem;padding-top:.7rem;border-top:3px solid var(--warm);font-size:clamp(1.45rem,3vw,2rem)}}h3{{color:var(--accent)}}p,h1,h2,h3,h4,a,figcaption,code{{overflow-wrap:anywhere;word-break:break-word}}blockquote{{margin:1.4rem 0;padding:1rem 1.25rem;border-left:5px solid var(--accent);background:var(--soft)}}code{{padding:.14em .4em;border-radius:5px;background:#edf0eb}}figure{{margin:1.6rem 0 2.2rem;padding:clamp(8px,2vw,18px);border:1px solid var(--line);border-radius:16px;background:#fff;box-shadow:0 10px 30px rgba(28,62,48,.09)}}img{{display:block;max-width:100%;height:auto;margin:auto;border-radius:10px}}figcaption{{margin-top:.7rem;text-align:center;color:#60726b}}a{{color:#087553}}.table-wrap{{overflow-x:auto;margin:1.4rem 0;border:1px solid var(--line);border-radius:12px}}table{{width:100%;border-collapse:collapse;background:#fff}}th,td{{padding:.72rem .85rem;border-bottom:1px solid var(--line);text-align:left;vertical-align:top}}th{{background:var(--soft)}}@media(max-width:640px){{main{{width:100%;margin:0;padding:24px 18px 48px;border:0;border-radius:0}}}}@media print{{html,body{{background:#fff}}main{{width:100%;margin:0;padding:0;border:0;box-shadow:none}}figure{{break-inside:avoid}}}}</style></head><body><main>{body}</main></body></html>'''

gfa_core/mas/agents/conversation_tools.py

@@ -23,10 +23,7 @@ from gfa_core.mas.mutation_surface import ToolSurfaceEntry
 from gfa_core.mas.stage_capabilities import get_stage_capability
 from gfa_core.mas.tools.stage_status import SUCCESS_STATUSES
 from gfa_core.mas.tools.stage_tools import stage_registry
-from gfa_core.mas.workflow_catalog import (
-    load_canonical_workflow_contract,
-    load_rnaseq_business_context,
-)
+from gfa_core.mas.workflow_catalog import load_rnaseq_business_context
 from gfa_core.mas.workflow_models import GoalLedger
 
 
@@ -56,6 +53,7 @@ class ConversationToolRegistry:
         *,
         workflow_id: str | None = None,
         guidance: bool = False,
+        language: str = "zh",
     ) -> str:
         text = str(query or "").strip()
         if self._analysis_handoff_resolver is None:
@@ -65,6 +63,7 @@ class ConversationToolRegistry:
                 text,
                 workflow_id=workflow_id,
                 guidance=guidance,
+                language=language,
             )
             or ""
         ).strip()
@@ -267,6 +266,7 @@ class RecognizeAnalysisIntentArgs(BaseModel):
     query: str
     job_id: str = ""
     prefix: str = ""
+    language: str = "zh"
 
     _job_id_is_local = field_validator("job_id")(_validate_job_id)
     _prefix_is_literal = field_validator("prefix")(_validate_prefix)
@@ -1607,9 +1607,11 @@ def build_plantai_conversation_tools(
         *,
         workflow_id: str | None = None,
         guidance: bool = False,
+        language: str = "zh",
     ) -> str:
         """Adapt MainAgent memory to the original IntentAgent text contract."""
         text = str(query or "").strip()
+        en = language == "en"
         if (
             text.startswith("执行数据准备（genome_prepare）")
             and "当前业务动作已由数据准备契约唯一确定" in text
@@ -1619,15 +1621,15 @@ def build_plantai_conversation_tools(
             return text
         terms = goal_ledger.user_terms
         context: list[str] = []
-        canonical_request = ""
         family = str(terms.get("gene_family") or "").strip()
         if family:
-            context.append(f"基因家族 {family}")
-        for key, label in (
-            ("target_species", "目标物种"),
-            ("reference_species", "参考物种"),
-            ("comparative_species", "比较物种"),
+            context.append(f"Gene family: {family}" if en else f"基因家族 {family}")
+        for key, label_zh, label_en in (
+            ("target_species", "目标物种", "Target species"),
+            ("reference_species", "参考物种", "Reference species"),
+            ("comparative_species", "比较物种", "Comparative species"),
         ):
+            label = label_en if en else label_zh
             values = terms.get(key) or ()
             if isinstance(values, (list, tuple)):
                 cleaned = [
@@ -1647,21 +1649,39 @@ def build_plantai_conversation_tools(
             if mode == "whole_genome":
                 enabled_designs = "、".join(business_context["enabled_designs"])
                 disabled_designs = "、".join(business_context["disabled_designs"])
-                context.append(
-                    f"RNA-seq业务模式 全基因组；{experiment_design} 是实验设计权威输入；"
-                    f"可解释设计为 {enabled_designs}，{disabled_designs} 本轮禁用；"
-                    "设计或 batch 降级不得静默发生"
-                )
+                if en:
+                    context.append(
+                        f"RNA-seq mode: whole genome; {experiment_design} is the authoritative experimental design input; "
+                        f"interpretable designs: {enabled_designs}, {disabled_designs} disabled this round; "
+                        "design or batch downgrades must not happen silently"
+                    )
+                else:
+                    context.append(
+                        f"RNA-seq业务模式 全基因组；{experiment_design} 是实验设计权威输入；"
+                        f"可解释设计为 {enabled_designs}，{disabled_designs} 本轮禁用；"
+                        "设计或 batch 降级不得静默发生"
+                    )
             else:
-                context.append(
-                    "RNA-seq业务模式 基因家族成员；使用已成功的全基因组差异表达结果"
-                    "和精确确认的基因家族成员进行投影；不得自动选择首个成员文件，"
-                    "不得隐式重跑差异表达"
-                )
+                if en:
+                    context.append(
+                        "RNA-seq mode: gene family member; use successful whole-genome differential expression results "
+                        "and precisely confirmed gene family members for projection; must not auto-select the first member file, "
+                        "must not implicitly re-run differential expression"
+                    )
+                else:
+                    context.append(
+                        "RNA-seq业务模式 基因家族成员；使用已成功的全基因组差异表达结果"
+                        "和精确确认的基因家族成员进行投影；不得自动选择首个成员文件，"
+                        "不得隐式重跑差异表达"
+                    )
         elif rna_route == "whole_genome":
-            context.append("RNA-seq模式 全基因组")
+            context.append(
+                "RNA-seq mode: whole genome" if en else "RNA-seq模式 全基因组"
+            )
         elif rna_route == "family_member":
-            context.append("RNA-seq模式 基因家族成员")
+            context.append(
+                "RNA-seq mode: gene family member" if en else "RNA-seq模式 基因家族成员"
+            )
         job_id = bound_job_id or str(terms.get("active_job_id") or "").strip()
         formal_config_exists = bool(
             job_id and (project_root / "input" / job_id / "config.xlsx").is_file()
@@ -1671,7 +1691,6 @@ def build_plantai_conversation_tools(
             or "config.xlsx 已确认" in goal_ledger.approved_decisions
             or formal_config_exists
         )
-        genome_prepare_observed = False
         subfamily_confirm_observed = False
         if job_id:
             final_dir = project_root / "workspace" / job_id / "final"
@@ -1682,8 +1701,6 @@ def build_plantai_conversation_tools(
                     continue
                 stage = str(manifest.get("stage") or "").strip()
                 status = manifest.get("status")
-                if stage == "genome_prepare" and status in {"success", "partial_success"}:
-                    genome_prepare_observed = True
                 manifest_prefix = str(manifest.get("prefix") or "").strip()
                 family_matches = not family or bool(
                     re.search(
@@ -1698,11 +1715,20 @@ def build_plantai_conversation_tools(
                     and family_matches
                 ):
                     subfamily_confirm_observed = True
+        genome_prepare_observed = False
+        canonical_request = None
         if config_confirmed and not genome_prepare_observed:
-            context.append(
-                "正式 config.xlsx 已确认；这只表示配置已核对，不等于数据准备已经完成，"
-                "当前 job 尚无数据准备完成证据；具体分析意图和阶段仍由现有分析系统判断"
-            )
+            if en:
+                context.append(
+                    "Formal config.xlsx confirmed; this indicates configuration has been verified, "
+                    "not that data preparation is complete. No data preparation completion evidence in current job; "
+                    "specific analysis intent and stages still determined by the existing analysis system"
+                )
+            else:
+                context.append(
+                    "正式 config.xlsx 已确认；这只表示配置已核对，不等于数据准备已经完成，"
+                    "当前 job 尚无数据准备完成证据；具体分析意图和阶段仍由现有分析系统判断"
+                )
             if guidance and workflow_id in {
                 "data_preparation",
                 "gene_family",
@@ -1717,19 +1743,34 @@ def build_plantai_conversation_tools(
                 )
                 actions = {item.id: item for item in contract.actions}
                 target_stage = actions[transition.action_id].target_stage
-                canonical_request = (
-                    f"执行数据准备（{target_stage}）\n"
-                    "当前业务动作已由数据准备契约唯一确定"
-                )
+                if en:
+                    canonical_request = (
+                        f"Execute data preparation ({target_stage})\n"
+                        "Business action uniquely determined by data preparation contract"
+                    )
+                else:
+                    canonical_request = (
+                        f"执行数据准备（{target_stage}）\n"
+                        "当前业务动作已由数据准备契约唯一确定"
+                    )
         if family and subfamily_confirm_observed:
-            context.append(
-                "当前 job 的 subfamily_confirm 已成功完成，家族成员和亚家族分类表已确认，"
-                "这是当前项目状态事实；具体分析意图和阶段仍由现有分析系统判断"
-            )
+            if en:
+                context.append(
+                    "Current job's subfamily_confirm completed successfully; "
+                    "family members and subfamily classification table confirmed. "
+                    "This is a current project state fact; specific analysis intent and stages still determined by the existing analysis system"
+                )
+            else:
+                context.append(
+                    "当前 job 的 subfamily_confirm 已成功完成，家族成员和亚家族分类表已确认，"
+                    "这是当前项目状态事实；具体分析意图和阶段仍由现有分析系统判断"
+                )
         if canonical_request:
             return canonical_request
         if not context:
             return text
+        if en:
+            return f"{text}\nConfirmed context: {'; '.join(context)}."
         return f"{text}\n已确认上下文：{'；'.join(context)}。"
 
     registry.set_analysis_handoff_resolver(intent_query_with_confirmed_context)
@@ -1746,7 +1787,7 @@ def build_plantai_conversation_tools(
             receiver="intent_v2",
             message_type="recognize_intent",
             payload={
-                "query": intent_query_with_confirmed_context(args.query),
+                "query": intent_query_with_confirmed_context(args.query, language=args.language),
                 "job_id": job_id,
                 "prefix": args.prefix,
                 "workspace": str(project_root / "workspace" / job_id) if job_id else "",

gfa_core/mas/agents/diagnosis.py

@@ -390,48 +390,57 @@ class DiagnosisAgent(BaseAgent):
     ERROR_PATTERNS = {
         "missing_input_file": {
             "patterns": [r"文件不存在", r"FileNotFoundError", r"No such file", r"找不到文件"],
-            "suggestion": "检查输入文件路径是否正确，确保文件已上传到input目录",
+            "suggestion_zh": "检查输入文件路径是否正确，确保文件已上传到input目录",
+            "suggestion_en": "Check if input file path is correct; ensure files are uploaded to the input directory",
             "auto_fix": False
         },
         "hmm_not_found": {
             "patterns": [r"required_pfams is empty", r"No HMM file", r"找不到HMM"],
-            "suggestion": "请上传HMM模型文件或指定--required-pfams参数",
+            "suggestion_zh": "请上传HMM模型文件或指定--required-pfams参数",
+            "suggestion_en": "Please upload HMM model file or specify --required-pfams parameter",
             "auto_fix": False
         },
         "blast_db_error": {
             "patterns": [r"BLAST Database error", r"makeblastdb failed"],
-            "suggestion": "Identify BLAST 固定用目标物种全蛋白 FASTA 建库；请优先检查目标全蛋白 FASTA 是否存在、非空且为有效蛋白FASTA格式，再检查参考家族蛋白 query 文件。",
+            "suggestion_zh": "Identify BLAST 固定用目标物种全蛋白 FASTA 建库；请优先检查目标全蛋白 FASTA 是否存在、非空且为有效蛋白FASTA格式，再检查参考家族蛋白 query 文件。",
+            "suggestion_en": "Identify BLAST uses target species full protein FASTA for database; check that target FASTA exists, is non-empty and valid, then check reference family protein query file.",
             "auto_fix": False
         },
         "memory_error": {
             "patterns": [r"MemoryError", r"out of memory", r"内存不足"],
-            "suggestion": "减少--cpu参数值或分批处理，建议使用--cpu 4",
+            "suggestion_zh": "减少--cpu参数值或分批处理，建议使用--cpu 4",
+            "suggestion_en": "Reduce --cpu parameter or use batch processing; recommend --cpu 4",
             "auto_fix": True,
             "fix_action": "reduce_cpu"
         },
         "rnaseq_library_strandness_error": {
             "patterns": [r"library_strandness", r"链特异性", r"rna-strandness", r"featureCounts.*-s"],
-            "suggestion": "请在 samples.yaml 的 rnaseq.library_strandness 或每个 samples.<sample>.library_strandness 中明确填写 non_stranded / forward / reverse；混合链特异性样本当前不能用单次 featureCounts 定量。",
+            "suggestion_zh": "请在 samples.yaml 的 rnaseq.library_strandness 或每个 samples.<sample>.library_strandness 中明确填写 non_stranded / forward / reverse；混合链特异性样本当前不能用单次 featureCounts 定量。",
+            "suggestion_en": "Please specify non_stranded / forward / reverse in samples.yaml rnaseq.library_strandness or each samples.<sample>.library_strandness; mixed strand-specific samples cannot be quantified in a single featureCounts run.",
             "auto_fix": False
         },
         "timeout_error": {
             "patterns": [r"Timeout", r"timed out", r"超时"],
-            "suggestion": "增加超时时间或检查输入数据大小",
+            "suggestion_zh": "增加超时时间或检查输入数据大小",
+            "suggestion_en": "Increase timeout or check input data size",
             "auto_fix": False
         },
         "conda_env_error": {
             "patterns": [r"conda", r"Environment", r"找不到命令"],
-            "suggestion": "检查conda环境是否正确激活，确保工具已安装",
+            "suggestion_zh": "检查conda环境是否正确激活，确保工具已安装",
+            "suggestion_en": "Check that conda environment is activated and tools are installed",
             "auto_fix": False
         },
         "permission_error": {
             "patterns": [r"Permission denied", r"权限不足"],
-            "suggestion": "检查文件权限，确保有读写权限",
+            "suggestion_zh": "检查文件权限，确保有读写权限",
+            "suggestion_en": "Check file permissions; ensure read/write access",
             "auto_fix": False
         },
         "invalid_parameter": {
             "patterns": [r"unrecognized arguments", r"invalid parameter", r"参数错误"],
-            "suggestion": "检查参数名称是否正确，注意使用短横线而非下划线",
+            "suggestion_zh": "检查参数名称是否正确，注意使用短横线而非下划线",
+            "suggestion_en": "Check parameter name; use hyphens instead of underscores",
             "auto_fix": True,
             "fix_action": "fix_parameter_format"
         },
@@ -448,7 +457,8 @@ class DiagnosisAgent(BaseAgent):
                 r"RuntimeError: MAFFT失败.*No such file",
                 r"RuntimeError:.*失败.*No such file or directory"
             ],
-            "suggestion": "生物信息学工具未安装或未添加到PATH。建议：1) 激活conda环境: conda activate test4py311 2) 安装缺失工具: conda install -c bioconda mafft blast iqtree hmmer",
+            "suggestion_zh": "生物信息学工具未安装或未添加到PATH。建议：1) 激活conda环境: conda activate test4py311 2) 安装缺失工具: conda install -c bioconda mafft blast iqtree hmmer",
+            "suggestion_en": "Bioinformatics tools not installed or not in PATH. Suggestions: 1) activate conda environment: conda activate test4py311 2) install missing tools: conda install -c bioconda mafft blast iqtree hmmer",
             "auto_fix": False,
             "severity": "high"
         },
@@ -461,7 +471,8 @@ class DiagnosisAgent(BaseAgent):
                 r"requires.*stage.*not completed",
                 r"dependency.*not satisfied"
             ],
-            "suggestion": "前置分析阶段未完成，系统将自动先执行依赖阶段",
+            "suggestion_zh": "前置分析阶段未完成，系统将自动先执行依赖阶段",
+            "suggestion_en": "Prerequisite analysis stage not completed; system will execute dependency stage first automatically",
             "auto_fix": True,
             "fix_action": "execute_dependencies",
             "severity": "high"
@@ -601,6 +612,7 @@ class DiagnosisAgent(BaseAgent):
         prepare_patch_workspace: bool = False,
         scope_id: str = "",
         params: Optional[Dict[str, Any]] = None,
+        language: str = "zh",
     ) -> Dict[str, Any]:
         """只读审计某个 job/stage 的产物、manifest、日志和输入配置。
 
@@ -930,6 +942,7 @@ class DiagnosisAgent(BaseAgent):
         gate: str = "preflight",
         project_root: Path = PROJECT_ROOT,
         params: Optional[Dict[str, Any]] = None,
+        language: str = "zh",
     ) -> Dict[str, Any]:
         """Global QA gate for stage preflight/postflight/failure.
 
@@ -4490,10 +4503,10 @@ class DiagnosisAgent(BaseAgent):
         except Exception as exc:
             print(f"[DiagnosisAgent] Failed to remember LLM diagnosis case: {exc}")
     
-    def _match_error_pattern(self, error_text: str) -> Optional[Dict]:
+    def _match_error_pattern(self, error_text: str, language: str = "zh") -> Optional[Dict]:
         """匹配错误模式"""
         error_lower = error_text.lower()
-        
+
         priority_order = [
             "tool_not_found",
             "dependency_failed",
@@ -4507,15 +4520,16 @@ class DiagnosisAgent(BaseAgent):
             "permission_error",
             "timeout_error"
         ]
-        
+
         for error_type in priority_order:
             if error_type in self.ERROR_PATTERNS:
                 pattern_info = self.ERROR_PATTERNS[error_type]
                 for pattern in pattern_info["patterns"]:
                     if re.search(pattern, error_text, re.IGNORECASE):
+                        suggestion_key = "suggestion_en" if language == "en" else "suggestion_zh"
                         return {
                             "error_type": error_type,
-                            "suggestion": pattern_info["suggestion"],
+                            "suggestion": pattern_info.get(suggestion_key, pattern_info.get("suggestion_zh", "")),
                             "auto_fix": pattern_info.get("auto_fix", False),
                             "fix_action": pattern_info.get("fix_action"),
                             "severity": pattern_info.get("severity", "medium")

gfa_core/mas/agents/intent_v2.py

@@ -4,7 +4,7 @@ IntentAgent V2 - LLM驱动的智能意图识别
 """
 import json
 import re
-from typing import Dict, List, Optional, Any
+from typing import Any, Dict, List, Mapping, Optional
 from .base import AgentMessage, AgentState, BaseAgent, QueryMode
 from gfa_core.mas.llm import LLMManager
 from gfa_core.mas.stage_catalog import (
@@ -152,6 +152,82 @@ class IntentAgentV2(BaseAgent):
     
     def __init__(self, llm_manager: LLMManager):
         super().__init__(llm_manager)
+
+    @staticmethod
+    def _apply_structured_goal_context(
+        intent: Dict[str, Any],
+        goal_context: Mapping[str, Any] | None,
+    ) -> Dict[str, Any]:
+        if not isinstance(goal_context, Mapping):
+            return intent
+        terms = goal_context.get("user_terms")
+        if not isinstance(terms, Mapping):
+            return intent
+
+        def species_list(value: Any) -> List[str]:
+            if isinstance(value, str):
+                values = value.split(",")
+            elif isinstance(value, (list, tuple, set)):
+                values = value
+            else:
+                values = ()
+            return list(
+                dict.fromkeys(str(item).strip() for item in values if str(item).strip())
+            )
+
+        normalized = dict(intent)
+        params = dict(normalized.get("params") or {})
+        targets = species_list(
+            params.get("targetSpecies") or normalized.get("target_species")
+        ) or species_list(terms.get("target_species"))
+        references = species_list(
+            params.get("referenceSpecies") or normalized.get("reference_species")
+        ) or species_list(terms.get("reference_species"))
+        comparatives = species_list(
+            params.get("comparativeSpecies") or normalized.get("comparative_species")
+        ) or species_list(terms.get("comparative_species"))
+        references = [species for species in references if species not in targets]
+        comparatives = [
+            species
+            for species in comparatives
+            if species not in targets and species not in references
+        ]
+        family = str(
+            normalized.get("gene_family") or terms.get("gene_family") or ""
+        ).strip()
+        scope_mode = (
+            "target_plus_reference_plus_comparative"
+            if references and comparatives
+            else "target_plus_reference"
+            if references
+            else "target_plus_comparative"
+            if comparatives
+            else "target_only"
+        )
+        params.update(
+            {
+                "targetSpecies": targets,
+                "referenceSpecies": references,
+                "comparativeSpecies": comparatives,
+                "target_species": targets,
+                "scopeMode": scope_mode,
+            }
+        )
+        normalized.update(
+            {
+                "params": params,
+                "gene_family": family,
+                "target_species": targets,
+                "reference_species": references,
+                "comparative_species": comparatives,
+                "scopeMode": scope_mode,
+            }
+        )
+        if targets:
+            normalized["species"] = targets[0] if len(targets) == 1 else "Multi"
+        if normalized.get("species") and family:
+            normalized["prefix"] = f"{normalized['species']}.{family}"
+        return normalized
     
     async def execute(self, state: AgentState, message: AgentMessage) -> AgentState:
         """
@@ -237,6 +313,11 @@ class IntentAgentV2(BaseAgent):
                     print(f"[IntentAgentV2] Protected species from parser: {base_intent['species']}")
             intent_result.setdefault("task_kind", "pipeline")
             intent_result.setdefault("taskType", intent_result.get("task_kind", "pipeline"))
+
+        intent_result = self._apply_structured_goal_context(
+            intent_result,
+            message.payload.get("goal_context"),
+        )
         
         # 保留原始 stages，不过滤（留给 PlanningAgentV2 处理）
         original_stages = intent_result.get("stages", [])
@@ -693,8 +774,17 @@ class IntentAgentV2(BaseAgent):
                 intent_data["stages"] = (
                     ["rnaseq_de", "rnaseq_go_enrich", "rnaseq_kegg_enrich", "rnaseq_fgsea", "rnaseq_wgcna"]
                     if has_rnaseq_keyword
-                    else ["identify"]
+                    else []
                 )
+                if not intent_data["stages"]:
+                    intent_data.update(
+                        {
+                            "description": "未能可靠识别要执行的分析阶段",
+                            "needs_clarification": True,
+                            "clarification_type": "analysis_stage",
+                            "clarification_question": "请说明要执行的具体分析阶段",
+                        }
+                    )
         
         # 根据依赖关系补充缺失的阶段（关键修复：确保依赖展开）
         original_stages = intent_data["stages"]
@@ -975,11 +1065,8 @@ class IntentAgentV2(BaseAgent):
             has_rnaseq_keyword = any(
                 kw in query_lower for kw in ["rnaseq", "rna-seq", "rna seq", "转录组", "fastq"]
             )
-            stages = (
-                ["rnaseq_de", "rnaseq_go_enrich", "rnaseq_kegg_enrich", "rnaseq_fgsea", "rnaseq_wgcna"]
-                if has_rnaseq_keyword
-                else ["identify"]
-            )
+            if has_rnaseq_keyword:
+                stages = ["rnaseq_de", "rnaseq_go_enrich", "rnaseq_kegg_enrich", "rnaseq_fgsea", "rnaseq_wgcna"]
         
         # 补充依赖（使用 stage_catalog 的主规划依赖链）
         original_stages = stages.copy()
@@ -989,18 +1076,31 @@ class IntentAgentV2(BaseAgent):
         gene_family = self._extract_gene_family_fallback(query)
         species = "Multi" if any(kw in query.lower() for kw in ["基因家族成员鉴定", "基因家族鉴定", "identify"]) else ""
         
-        return {
+        result = {
             "gene_family": gene_family,
             "species": species,
             "prefix": f"{species}.{gene_family}",
             "stages": stages,
-            "description": f"执行{gene_family}基因家族的{', '.join(stages)}分析",
+            "description": (
+                f"执行{gene_family}基因家族的{', '.join(stages)}分析"
+                if stages
+                else "未能可靠识别要执行的分析阶段"
+            ),
             "requirements": [],
-            "confidence": 0.7,
+            "confidence": 0.7 if stages else 0.0,
             "task_kind": "pipeline",
             "taskType": "pipeline",
             "mode": "fallback"
         }
+        if not stages:
+            result.update(
+                {
+                    "needs_clarification": True,
+                    "clarification_type": "analysis_stage",
+                    "clarification_question": "请说明要执行的具体分析阶段",
+                }
+            )
+        return result
 
 
 def create_intent_agent_v2(llm_manager):

gfa_core/mas/agents/interpretation_evidence.py

@@ -193,14 +193,24 @@ def _read_json(path: Path) -> Dict[str, Any]:
 
 def _artifact_path(manifest: Dict[str, Any], artifact_key: str) -> Optional[Path]:
     artifact = (manifest.get("artifacts") or {}).get(artifact_key) or {}
-    path_text = artifact.get("full_path") or artifact.get("path")
-    if path_text:
+    manifest_path = Path(str(manifest.get("_manifest_path") or ""))
+    job_dir = manifest_path.parent.parent if manifest_path else None
+    for path_text in (artifact.get("full_path"), artifact.get("path")):
+        if not path_text:
+            continue
         path = Path(str(path_text))
-        if not path.is_absolute() and manifest.get("_manifest_path"):
-            job_dir = Path(str(manifest["_manifest_path"])).parent.parent
-            path = job_dir / path
-        if path.exists():
-            return path
+        candidates = [path]
+        if job_dir:
+            if not path.is_absolute():
+                candidates.append(job_dir / path)
+            else:
+                job_index = next((index for index, part in enumerate(path.parts) if part.startswith("job_")), None)
+                if job_index is not None:
+                    candidates.append(job_dir.joinpath(*path.parts[job_index + 1 :]))
+                candidates.append(manifest_path.parent / path.name)
+        for candidate in candidates:
+            if candidate.exists():
+                return candidate
     return None
 
 

gfa_core/mas/agents/router.py

@@ -54,6 +54,7 @@ class RouterAgent(BaseAgent):
             "history": history,
             "session_id": str(payload.get("session_id") or state.session_id),
             "job_id": str(payload.get("job_id") or state.job_id),
+            "language": str(payload.get("language") or "zh"),
         }
         if "goal_ledger" in payload:
             runtime_context["goal_ledger"] = payload.get("goal_ledger")

gfa_core/mas/context_index.py

@@ -251,7 +251,13 @@ class ProjectContextIndexer:
         self.workspace_root = self.project_root / "workspace"
         self.allow_legacy_manifest_migration = allow_legacy_manifest_migration
 
-    def build(self, job_id: str, prefix: str = "") -> ProjectContextIndex:
+    def build(
+        self,
+        job_id: str,
+        prefix: str = "",
+        *,
+        include_stage_status: bool = True,
+    ) -> ProjectContextIndex:
         workspace = self.workspace_root / job_id
         canonical_input_dir = self.project_root / "input" / job_id
         input_dir = canonical_input_dir if canonical_input_dir.exists() else workspace / "input"
@@ -275,7 +281,8 @@ class ProjectContextIndexer:
         self._scan_files(index, final_dir, "final")
         self._scan_stage_dirs(index, workspace)
         self._load_manifests(index, final_dir)
-        self._load_stage_status(index, job_id, prefix or self._best_prefix(index))
+        if include_stage_status:
+            self._load_stage_status(index, job_id, prefix or self._best_prefix(index))
         self._scan_logs(index, logs_dir)
         self._derive_warnings(index)
         return index
@@ -696,11 +703,16 @@ def build_project_context(
     project_root: Path | str = PROJECT_ROOT,
     *,
     allow_legacy_manifest_migration: bool = True,
+    include_stage_status: bool = True,
 ) -> ProjectContextIndex:
     return ProjectContextIndexer(
         project_root,
         allow_legacy_manifest_migration=allow_legacy_manifest_migration,
-    ).build(job_id=job_id, prefix=prefix)
+    ).build(
+        job_id=job_id,
+        prefix=prefix,
+        include_stage_status=include_stage_status,
+    )
 
 
 def main() -> int:

gfa_core/mas/core/fast_intent.py

@@ -4,7 +4,7 @@
 import re
 from typing import Optional, Dict, Any, List
 from gfa_core.mas.stage_catalog import complete_dependencies as catalog_complete_dependencies
-from gfa_core.mas.utils.intent_parser import is_gwas_fmt_intent, parse_gene_family_scope_selection, resolve_gwas_intent, resolve_population_intent
+from gfa_core.mas.utils.intent_parser import is_gwas_fmt_intent, parse_gene_family_scope_selection, resolve_gene_family_report_intent, resolve_gwas_intent, resolve_population_intent
 
 # 分析类型关键词映射（注意：key必须与orchestrator.py中的analysis_type匹配）
 ANALYSIS_KEYWORDS = {
@@ -96,6 +96,12 @@ def fast_intent_recognition(query: str) -> Optional[Dict[str, Any]]:
         None: 无法识别
     """
     query_lower = query.lower()
+    report_intent = resolve_gene_family_report_intent(query)
+    if report_intent:
+        return {
+            "stage": "gene_family_report",
+            **report_intent,
+        }
     population = resolve_population_intent(query)
     if population:
         return {

gfa_core/mas/core/orchestrator.py

@@ -29,6 +29,74 @@ WORKSPACE_DIR = WORKSPACE_ROOT
 
 
 class Orchestrator:
+    # 双语消息映射表：中文 → English（用于用户可见消息）
+    _MSG = {
+        # analysis_node — response 消息
+        "✅ 所有分析阶段已完成！": "✅ All analysis stages completed!",
+        "📋 自动规划剩余分析阶段": "📋 Auto-planned remaining analysis stages",
+        "**已完成**": "**Completed**",
+        "无": "None",
+        "**待执行**": "**Pending**",
+        "个阶段)": "stage(s))",
+        "请确认是否按顺序执行以上分析阶段？": "Confirm to execute the above analysis stages in order?",
+        "我将帮您执行RNA-seq": "I will help you execute RNA-seq",
+        "分析。请确认参数后开始。": "analysis. Please confirm parameters to start.",
+        "RNA-seq分析将通过前端异步执行。": "RNA-seq analysis will be executed asynchronously via frontend.",
+        " 分析。PlanningAgentV2 已根据当前状态生成完整执行计划。": " analysis. PlanningAgentV2 has generated a complete execution plan based on current state.",
+        "我将帮您执行": "I will help you execute",
+        "分析。": "analysis.",
+        "请提供基因家族蛋白序列文件（.fa/.fasta），用于MEME保守基序预测。": "Please provide gene family protein sequence file (.fa/.fasta) for MEME conserved motif prediction.",
+        "请提供config.xlsx及基因组/GFF/CDS/PEP等原始文件，用于完整数据准备。": "Please provide config.xlsx and genome/GFF/CDS/PEP files for complete data preparation.",
+        "请提供基因家族蛋白序列文件，用于Pfam/InterPro结构域注释。": "Please provide gene family protein sequence file for Pfam/InterPro domain annotation.",
+        "请提供基因家族的GFF注释文件和基因组序列。": "Please provide gene family GFF annotation file and genome sequence.",
+        "请提供基因家族成员的启动子序列（上游2000bp）或全基因组序列+GFF。": "Please provide gene family promoter sequences (upstream 2000bp) or whole genome + GFF.",
+        "请提供物种基因组序列（.fasta）和GFF注释文件。": "Please provide species genome sequence (.fasta) and GFF annotation file.",
+        "请提供多个物种的基因组序列和GFF注释文件，以及物种间的同源关系。": "Please provide genome sequences and GFF annotation files for multiple species, along with inter-species orthology.",
+        "请提供染色体长度信息和基因位置信息。": "Please provide chromosome length and gene position information.",
+        "请提供拟南芥同源基因映射文件（ortholog_pairs.txt，格式：拟南芥ID\t苦荞ID），用于STRING数据库互作预测。如未提供，请先运行同源基因识别分析。": "Please provide Arabidopsis ortholog mapping file (ortholog_pairs.txt, format: AtID\tTargetID) for STRING interaction prediction. If unavailable, run ortholog identification first.",
+        "请提供目标物种和参考物种（如拟南芥At）的蛋白序列文件，用于同源基因识别。": "Please provide target and reference species (e.g., Arabidopsis At) protein FASTA for ortholog identification.",
+        "请提供物种间同源基因对（ortholog pairs，格式：基因A\t基因B）或共线性分析结果。": "Please provide inter-species ortholog pairs (format: GeneA\tGeneB) or synteny analysis results.",
+        "请提供RNA-seq表达矩阵（FPKM/TPM/Count）或原始测序数据。": "Please provide RNA-seq expression matrix (FPKM/TPM/Count) or raw sequencing data.",
+        "请提供差异表达基因列表（DEGs）或目标基因集。": "Please provide differentially expressed gene list (DEGs) or target gene set.",
+        "请提供基因家族蛋白序列文件，用于理化性质分析。": "Please provide gene family protein sequence file for physicochemical property analysis.",
+        "请提供基因家族HMM文件和蛋白序列文件，将自动执行标准11阶段分析。": "Please provide gene family HMM file and protein FASTA; standard 11-stage analysis will be executed automatically.",
+        "请提供RNA-seq原始测序数据（FASTQ格式）。": "Please provide RNA-seq raw sequencing data (FASTQ format).",
+        "请提供质控后的clean reads（FASTQ格式）和参考基因组。": "Please provide QC-filtered clean reads (FASTQ format) and reference genome.",
+        "请提供比对后的BAM文件或参考基因组注释GFF。": "Please provide aligned BAM file or reference genome annotation GFF.",
+        "请提供featureCounts计数矩阵和样本分组表，用于VST、PCA、样本相关性和聚类。": "Please provide featureCounts count matrix and sample grouping table for VST, PCA, sample correlation and clustering.",
+        "请提供多时相表达矩阵（如不同发育阶段/处理时间点）。": "Please provide multi-temporal expression matrix (e.g., different developmental stages/treatment time points).",
+        "请提供VST/log2表达矩阵或count矩阵和样本性状表，用于共表达网络构建。": "Please provide VST/log2 expression matrix or count matrix and sample trait table for co-expression network construction.",
+        "请提供差异表达结果和gene2go背景文件，或先执行背景注释。": "Please provide differential expression results and gene2go background file, or run background annotation first.",
+        "请提供差异表达结果和gene2ko背景文件，或先执行背景注释。": "Please provide differential expression results and gene2ko background file, or run background annotation first.",
+        "请提供差异表达排序统计量和gene2ko/KEGG背景，用于fgsea排序富集。": "Please provide differential expression ranking statistics and gene2ko/KEGG background for fgsea ranked enrichment.",
+        "请提供差异表达结果和目标物种蛋白FASTA；若已有InterProScan TSV会优先复用。": "Please provide differential expression results and target species protein FASTA; existing InterProScan TSV will be reused if available.",
+        "请提供表达矩阵（样本数≥6），用于共表达网络构建。": "Please provide expression matrix (samples ≥ 6) for co-expression network construction.",
+        # knowledge_node — system prompt / response 消息
+        "你是BotanicalBaseAgent基因家族分析系统的知识助手。基于提供的文献资料，生成专业、详细的回答。": "You are a knowledge assistant for the BotanicalBaseAgent gene family analysis system. Generate professional, detailed answers based on provided literature.",
+        "📚 基于知识库的回答：": "📚 Knowledge base answer:",
+        "**参考文献**：": "**References**:",
+        "（来源: 本地LLM回答 + RAG辅助检索）": "(Source: local LLM answer + RAG-assisted retrieval)",
+        "⚠️ (本地LLM回答，未使用RAG文献检索)": "⚠️ (local LLM answer, RAG retrieval not used)",
+        "找到": "Found",
+        "条相关文献，正在生成本地LLM回答...": " related references; generating local LLM answer...",
+        "未找到相关文献，生成一般回答": "No related references found; generating general answer",
+        "你是一个专业的基因家族分析意图识别器。": "You are a professional gene family analysis intent recognizer.",
+        "你是一个基因家族分析专家。用户想要进行基因家族分析。": "You are a gene family analysis expert. The user wants to perform gene family analysis.",
+    }
+
+    @classmethod
+    def _l(cls, zh_text: str, language: str = "zh") -> str:
+        """Return language-appropriate text from the bilingual map."""
+        if language == "zh":
+            return zh_text
+        en = cls._MSG.get(zh_text)
+        if en is not None:
+            return en
+        # fallback: try partial match for template strings
+        for zh_key, en_val in cls._MSG.items():
+            if zh_key in zh_text:
+                return zh_text.replace(zh_key, en_val)
+        return zh_text
     def __init__(
         self,
         llm_manager: LLMManager,
@@ -313,6 +381,7 @@ class Orchestrator:
 
     async def _router_node(self, state: GraphState) -> dict:
         query = state["query"]
+        language = state.get("language", "zh")
         
         # 使用 RouterAgent 进行意图路由
         print(f"[RouterNode] Using RouterAgent to classify query: {query[:50]}...")
@@ -378,6 +447,7 @@ class Orchestrator:
 
     async def _knowledge_node(self, state: GraphState) -> dict:
         query = state["query"]
+        language = state.get("language", "zh")
 
         try:
             # 优先调用当前 Dual-KB QA 服务
@@ -578,7 +648,7 @@ class Orchestrator:
                 "status": "completed",
                 "mode": "analysis",
                 "final_output": {
-                    "response": f"✅ 所有分析阶段已完成！\n\n已完成: {', '.join(completed_stages)}",
+                    "response": self._l("✅ 所有分析阶段已完成！", language) + "\n\n" + self._l("**已完成**", language) + ": " + ", ".join(completed_stages),
                     "analysis_plan": {
                         "gene_family": gene_family,
                         "species": species,
@@ -595,14 +665,7 @@ class Orchestrator:
             "status": "completed",
             "mode": "analysis",
             "final_output": {
-                "response": f"""📋 自动规划剩余分析阶段
-
-**已完成**: {', '.join(completed_stages) if completed_stages else '无'}
-
-**待执行** ({len(remaining_stages)}个阶段):
-{stages_text}
-
-请确认是否按顺序执行以上分析阶段？""",
+                "response": f"{self._l('📋 自动规划剩余分析阶段', language)}\n\n{self._l('**已完成**', language)}: {', '.join(completed_stages) if completed_stages else self._l('无', language)}\n\n{self._l('**待执行**', language)} ({len(remaining_stages)}{self._l('个阶段)', language)}:\n{stages_text}\n\n{self._l('请确认是否按顺序执行以上分析阶段？', language)}",
                 "analysis_plan": {
                     "gene_family": gene_family,
                     "species": species,
@@ -630,6 +693,7 @@ class Orchestrator:
     async def _analysis_node(self, state: GraphState) -> dict:
         query = state["query"]
         files = state.get("files", [])
+        language = state.get("language", "zh")
         print(f"[_analysis_node] Entered with query: {query[:50]}...", flush=True)
 
         try:
@@ -821,7 +885,7 @@ class Orchestrator:
                         "status": "completed",
                         "mode": "analysis",
                         "final_output": {
-                            "response": f"我将帮您执行RNA-seq {analysis_type}分析。请确认参数后开始。",
+                            "response": self._l("我将帮您执行RNA-seq", language) + " " + analysis_type + self._l("分析。请确认参数后开始。", language),
                             "analysis_plan": {
                                 "gene_family": gene_family,
                                 "species": species,
@@ -831,7 +895,7 @@ class Orchestrator:
                                     "prefix": f"{species}.{gene_family}",
                                     "cpu": 8
                                 },
-                                "guidance": "RNA-seq分析将通过前端异步执行。"
+                                "guidance": self._l("RNA-seq分析将通过前端异步执行。", language)
                             }
                         }
                     }
@@ -878,7 +942,7 @@ class Orchestrator:
                             "status": "completed",
                             "mode": "analysis",
                             "final_output": {
-                                "response": f"我将帮您执行RNA-seq {analysis_type} 分析。PlanningAgentV2 已根据当前状态生成完整执行计划。",
+                                "response": self._l("我将帮您执行RNA-seq", language) + " " + analysis_type + self._l(" 分析。PlanningAgentV2 已根据当前状态生成完整执行计划。", language),
                                 "analysis_plan": plan_result.get("final_output", {}).get("analysis_plan", {})
                             }
                         }
@@ -889,7 +953,7 @@ class Orchestrator:
                             "status": "completed",
                             "mode": "analysis",
                             "final_output": {
-                                "response": f"我将帮您执行RNA-seq {analysis_type} 分析。请确认参数后开始。",
+                                "response": self._l("我将帮您执行RNA-seq", language) + " " + analysis_type + self._l(" 分析。请确认参数后开始。", language),
                                 "analysis_plan": {
                                     "gene_family": gene_family,
                                     "species": species,
@@ -900,21 +964,25 @@ class Orchestrator:
                                         "prefix": f"{species}.{gene_family}",
                                         "cpu": 8
                                     },
-                                    "guidance": f"RNA-seq {analysis_type} 分析将通过前端异步执行。"
+                                    "guidance": self._l("RNA-seq " + analysis_type + " 分析将通过前端异步执行。", language)
                                 }
                             }
                         }
                 
                 # Ka/Ks分析：走通用规划流程（返回analysis_plan让前端执行）
                 # 注意：不再直接调用_execute_kaks，而是通过PlanningAgentV2规划完整流程
-                guidance = guidance_map.get(analysis_type, f"请上传进行{analysis_type}分析所需的文件，或确认使用系统内置数据。")
-                
+                guidance = guidance_map.get(analysis_type)
+                if guidance:
+                    guidance = self._l(guidance, language)
+                else:
+                    guidance = self._l("请上传进行", language) + analysis_type + self._l("分析所需的文件，或确认使用系统内置数据。", language)
+
                 return {
                     "current_agent": "analysis",
                     "status": "completed",
                     "mode": "analysis",
                     "final_output": {
-                        "response": f"我将帮您执行{analysis_type}分析。",
+                        "response": self._l("我将帮您执行", language) + analysis_type + self._l("分析。", language),
                         "analysis_plan": {
                             "gene_family": gene_family,
                             "species": species,

gfa_core/mas/llm/providers/deepseek.py

@@ -62,6 +62,9 @@ class DeepSeekProvider(BaseLLMProvider):
         tool_choice = kwargs.get("tool_choice")
         if tool_choice is not None:
             request_kwargs["tool_choice"] = tool_choice
+        response_format = kwargs.get("response_format")
+        if response_format is not None:
+            request_kwargs["response_format"] = response_format
 
         response = self.client.chat.completions.create(**request_kwargs)
         message = response.choices[0].message

gfa_core/mas/persistent_conversation.py

@@ -9,7 +9,7 @@ from typing import Any
 from uuid import NAMESPACE_URL, UUID, uuid4, uuid5
 
 from gfa_core.mas.agent_state_repository import SQLiteAgentStateRepository
-from gfa_core.mas.workflow_models import GoalLedger
+from gfa_core.mas.workflow_models import CurrentRecommendation, GoalLedger
 
 
 _INTERACTION_KINDS = {
@@ -68,7 +68,7 @@ def _history(items: Any) -> tuple[dict[str, str], ...]:
         content = str(item.get("content") or "")
         if content:
             normalized.append({"role": item["role"], "content": content})
-    return tuple(normalized[-10:])
+    return tuple(normalized)
 
 
 class PersistentConversationStore:
@@ -207,7 +207,7 @@ class PersistentConversationStore:
                 ),
                 expected_revision=ledger.revision,
             )
-        stored_messages = self.repository.list_recent_messages(conversation_id, limit=10)
+        stored_messages = self.repository.list_messages(conversation_id)
         if not stored_messages:
             for item in _history(client_history):
                 persisted = self.repository.append_message(
@@ -220,10 +220,7 @@ class PersistentConversationStore:
                 conversation = self.repository.get_conversation(conversation_id)
                 assert conversation is not None
                 assert conversation.revision == persisted.revision
-            stored_messages = self.repository.list_recent_messages(
-                conversation_id,
-                limit=10,
-            )
+            stored_messages = self.repository.list_messages(conversation_id)
         current_user_already_saved = bool(
             stored_messages
             and stored_messages[-1].role == "user"
@@ -264,6 +261,7 @@ class PersistentConversationStore:
         action: str | None,
         intent: dict[str, Any] | None = None,
         proposal: dict[str, str] | None = None,
+        recommendation: Mapping[str, Any] | None = None,
     ) -> GoalLedger:
         if interaction_kind not in _INTERACTION_KINDS:
             raise ValueError("interaction_kind is invalid")
@@ -362,6 +360,20 @@ class PersistentConversationStore:
         user_terms = dict(serialized["user_terms"])
         if turn.job_id:
             user_terms["active_job_id"] = turn.job_id
+        if action == "show_result":
+            if recommendation is None:
+                user_terms.pop("current_recommendation", None)
+            else:
+                current_recommendation = CurrentRecommendation.model_validate(
+                    recommendation
+                )
+                if current_recommendation.job_id != turn.job_id:
+                    raise ValueError("recommendation job binding conflict")
+                user_terms["current_recommendation"] = (
+                    current_recommendation.model_dump(mode="json", exclude_none=True)
+                )
+        if interaction_kind in {"cancel", "correction", "switch"}:
+            user_terms.pop("current_recommendation", None)
         if action == "clarify" and isinstance(proposal, dict):
             normalized_proposal = {
                 key: str(proposal.get(key) or "").strip()
@@ -381,6 +393,16 @@ class PersistentConversationStore:
                     "rnaseq_family_member",
                 }:
                     normalized_proposal["workflow_id"] = workflow_id
+                action_binding = {
+                    key: str(proposal.get(key) or "").strip()
+                    for key in (
+                        "workflow_action_id",
+                        "authorized_stages",
+                        "target_stage",
+                    )
+                }
+                if all(action_binding.values()):
+                    normalized_proposal.update(action_binding)
                 normalized_proposal["job_id"] = turn.job_id
                 user_terms["pending_proposal"] = normalized_proposal
         if (

gfa_core/mas/stage_capabilities.py

@@ -812,6 +812,23 @@ CAPABILITIES: Dict[str, StageCapability] = {
         repair_candidates=["manifest_recovery", "artifact_reresolve"],
         rerun_policy="auto_after_manifest_recovery",
     ),
+    "gene_family_report": StageCapability(
+        stage="gene_family_report",
+        cli="apps/cli/run_gene_family_report.py",
+        env="test4py311",
+        category="report",
+        consumes=[],
+        provides=[
+            ArtifactProduct("gene_family_report", "markdown", "{prefix}.gene_family_report.md"),
+            ArtifactProduct("gene_family_report_evidence", "json", "{prefix}.gene_family_report.evidence.json"),
+            ArtifactProduct("gene_family_report_model", "json", "{prefix}.gene_family_report.model.json"),
+            ArtifactProduct("gene_family_report_manifest", "manifest", "{prefix}.gene_family_report.manifest.json"),
+        ],
+        scope_modes=["target_only", "target_plus_reference", "all_selected"],
+        preflight_checks=["at_least_one_successful_planned_manifest"],
+        postflight_checks=["formal_report_four_files_exist"],
+        rerun_policy="open_reuses_fresh_generate_rebuilds",
+    ),
 }
 
 

gfa_core/mas/stage_catalog.py

@@ -78,6 +78,7 @@ PIPELINE_STAGES = {
     "ppi",
     "kaks",
     "combo",
+    "gene_family_report",
     "rnaseq_align",
     "rnaseq_quant",
     "rnaseq_sample_qc",
@@ -391,6 +392,7 @@ STAGE_ORDER = [
     "rnaseq_fgsea",
     "rnaseq_candidate_family",
     "rnaseq_enrich",
+    "gene_family_report",
 ]
 
 STAGE_INFO = {
@@ -441,6 +443,7 @@ STAGE_INFO = {
     "msa": {"name": "多序列比对", "estimated_time": 10, "cpu": 4},
     "enrichment": {"name": "富集分析", "estimated_time": 15, "cpu": 4},
     "combo": {"name": "组合图", "estimated_time": 10, "cpu": 4},
+    "gene_family_report": {"name": "基因家族结果报告", "estimated_time": 2, "cpu": 1},
     "genome_prepare": {"name": "基因组准备", "estimated_time": 10, "cpu": 4},
     "ncbi_fetch_sequences": {"name": "NCBI序列下载", "estimated_time": 5, "cpu": 4},
     "sequence_extract_by_id": {"name": "按ID提取序列", "estimated_time": 5, "cpu": 4},
@@ -529,6 +532,7 @@ EXECUTABLE_STAGES = {
     "ppi",
     "kaks",
     "combo",
+    "gene_family_report",
     "rnaseq_align",
     "rnaseq_quant",
     "rnaseq_sample_qc",

gfa_core/mas/stage_requirements.py

@@ -14,6 +14,7 @@ flag enables decision-making.
 """
 from __future__ import annotations
 
+from copy import copy
 from dataclasses import asdict, dataclass, field
 from pathlib import Path
 from typing import Any, Iterable
@@ -93,6 +94,7 @@ def resolve_stage_requirements(
     params: dict[str, Any] | None = None,
     project_root: Path | str = PROJECT_ROOT,
     shadow_only: bool = True,
+    project_context: ProjectContextIndex | None = None,
 ) -> StageRequirementReport:
     """Resolve stage requirements without changing planning or files."""
     params = dict(params or {})
@@ -110,7 +112,16 @@ def resolve_stage_requirements(
         report.messages.append("Cannot resolve requirements without job_id.")
         return report
 
-    context = build_project_context(job_id=job_id, prefix=identity.prefix, project_root=project_root)
+    context = project_context
+    if context is None:
+        context = build_project_context(
+            job_id=job_id,
+            prefix=identity.prefix,
+            project_root=project_root,
+            include_stage_status=False,
+        )
+    else:
+        context = copy(context)
     _apply_identity_to_context(context, identity)
 
     if identity.stage == "identify":
@@ -151,18 +162,28 @@ def resolve_many_stage_requirements(
     reports: dict[str, dict[str, Any]] = {}
     params_by_stage = params_by_stage or {}
     shared_params = shared_params or {}
+    contexts_by_prefix: dict[str, ProjectContextIndex] = {}
     for stage in stages:
         stage_params = {**shared_params, **(params_by_stage.get(stage) or {})}
         if stage not in CORE_REQUIREMENT_STAGES:
             continue
         try:
+            effective_prefix = stage_params.get("prefix") or prefix
+            if effective_prefix not in contexts_by_prefix:
+                contexts_by_prefix[effective_prefix] = build_project_context(
+                    job_id=job_id,
+                    prefix=effective_prefix,
+                    project_root=project_root,
+                    include_stage_status=False,
+                )
             report = resolve_stage_requirements(
                 stage=stage,
-                prefix=stage_params.get("prefix") or prefix,
+                prefix=effective_prefix,
                 job_id=job_id,
                 params=stage_params,
                 project_root=project_root,
                 shadow_only=shadow_only,
+                project_context=contexts_by_prefix[effective_prefix],
             )
             reports[stage] = report.to_dict()
         except Exception as exc:

gfa_core/mas/tools/stage_tools.py

@@ -559,9 +559,15 @@ class StageToolExecutor:
                     tfbs_top_motifs=(extra_params or {}).get("tfbsTopMotifs", 30),
                     tfbs_scan_both_strands=bool((extra_params or {}).get("tfbsScanBothStrands", True)),
                     scope_file=(extra_params or {}).get("scopeFile", "") or (extra_params or {}).get("scope_file", ""),
+                    scope_id=param_value("scopeId", "scope_id", default=""),
+                    rnaseq_scope_id=param_value("rnaseqScopeId", "rnaseq_scope_id", default=""),
                     analysis_route=(extra_params or {}).get("analysisRoute", "") or (extra_params or {}).get("analysis_route", ""),
                     member_set_id=(extra_params or {}).get("memberSetId", "") or (extra_params or {}).get("member_set_id", ""),
                     analysis_mode=(extra_params or {}).get("analysisMode", "") or (extra_params or {}).get("analysis_mode", ""),
+                    expected_stages=param_value("expectedStages", "expected_stages", default=[]),
+                    stage_statuses=param_value("stageStatuses", "stage_statuses", default={}),
+                    report_action=param_value("reportAction", "report_action", default="generate"),
+                    report_use_model=bool_param("reportUseModel", "report_use_model", default=True),
                     species_queue=(extra_params or {}).get("speciesQueue", []) or (extra_params or {}).get("species_queue", []),
                     target_species=target_species,
                     interspecies_ref_species=interspecies_ref_species,
@@ -1048,6 +1054,22 @@ class StageToolRegistry:
             "parameters": {"scope_aware": True},
             "timeout": 1800
         },
+        {
+            "id": "gene_family_report",
+            "name": "Gene Family Result Report",
+            "script": "run_gene_family_report.py",
+            "category": "report",
+            "description": "Generate or open the formal report for exactly the confirmed gene-family stages",
+            "required_inputs": [],
+            "outputs": [
+                "gene_family_report",
+                "gene_family_report_evidence",
+                "gene_family_report_model",
+                "gene_family_report_manifest",
+            ],
+            "parameters": {"action": "generate", "useModel": True},
+            "timeout": 1800
+        },
         # 15. Genome Prepare
         {
             "id": "genome_prepare",

gfa_core/mas/utils/intent_parser.py

@@ -87,6 +87,7 @@ STAGE_KEYWORDS = {
     "rnaseq_kegg_enrich": ["rnaseq kegg", "rnaseq_kegg", "rna kegg", "kegg富集", "kegg分析", "kegg", "富集分析", "enrich"],
     "rnaseq_enrich": ["富集分析", "enrich", "go分析", "kegg", "clusterprofiler"],
     "rnaseq_candidate_family": ["候选家族", "候选基因家族", "家族排名", "tf注释", "tf family", "tffamily", "pfam注释", "deg注释", "差异基因注释", "差异基因家族", "candidate family"],
+    "rnaseq_report": ["rnaseq结果报告", "rna-seq结果报告", "转录组结果报告", "rnaseq报告", "rna-seq报告"],
     "ncbi_fetch_sequences": ["ncbi下载", "ncbi序列", "下载序列", "下载蛋白", "获取蛋白", "下载protein", "下载cds", "protein和cds", "蛋白和cds", "基因list", "gene list", "gene_list", "id列表", ".ids"],
 }
 
@@ -256,8 +257,38 @@ RNASEQ_STAGES = {
     "rnaseq_kegg_enrich",
     "rnaseq_enrich",
     "rnaseq_candidate_family",
+    "rnaseq_report",
 }
 
+
+def resolve_gene_family_report_intent(query: str) -> Optional[Dict]:
+    """Resolve only explicit gene-family report generate/open requests."""
+    text = str(query or "").lower()
+    compact = re.sub(r"\s+", "", text)
+    if any(marker.replace(" ", "") in compact for marker in RNASEQ_MARKERS + ["转录组"]):
+        return None
+    action = ""
+    if any(phrase in compact for phrase in ("打开基因家族结果报告", "打开基因家族报告", "openresultreport")):
+        action = "open"
+    elif any(phrase in compact for phrase in ("生成基因家族结果报告", "生成基因家族报告", "生成结果报告", "generateresultreport")):
+        action = "generate"
+    if not action:
+        return None
+    return {
+        "gene_family": "",
+        "species": "",
+        "stages": ["gene_family_report"],
+        "primary_stage": "gene_family_report",
+        "task_type": "report",
+        "task_kind": "report",
+        "params": {"geneFamilyReportAction": action, "reportAction": action},
+        "confidence": 1.0,
+        "need_clarify": False,
+        "clarify_type": None,
+        "message": None,
+        "method": "gene_family_report_deterministic",
+    }
+
 # 物种关键词映射
 SPECIES_KEYWORDS = {
     "Ft": ["ft", "tartary", "苦荞", "荞麦"],
@@ -273,8 +304,21 @@ SPECIES_KEYWORDS = {
     "Cp": ["cp", "papaya", "木瓜"],
     "Vv": ["vv", "grape", "葡萄"],
     "Lg": ["lg"],
+    "Gm": ["gm"],
+    "Br": ["br"],
+    "Pt": ["pt"],
+    "Sb": ["sb"],
 }
 
+_SPECIES_CODE_PATTERN = (
+    r"(?<![A-Za-z0-9])(?:"
+    + "|".join(
+        sorted((re.escape(code) for code in SPECIES_KEYWORDS), key=len, reverse=True)
+    )
+    + r")(?![A-Za-z0-9])"
+)
+_ROLE_BOUND_SPECIES_CODE = r"[A-Z][A-Za-z0-9_-]{1,15}"
+
 SPECIES_TO_LATIN = {
     "At": "Arabidopsis thaliana",
     "Os": "Oryza sativa",
@@ -317,7 +361,11 @@ def _species_mentions_with_roles(query: str) -> Dict[str, List[str]]:
     def collect_species(segment: str) -> List[str]:
         segment_lower = segment.lower()
         found: List[str] = []
-        for match in re.finditer(r"(?<![A-Za-z0-9])(?:Ft|At|Os|Ls|Zm|Sl|Lc|Dl|Ppy|Py|Cp|Vv|Lg)(?![A-Za-z0-9])", segment, flags=re.IGNORECASE):
+        for match in re.finditer(
+            _SPECIES_CODE_PATTERN,
+            segment,
+            flags=re.IGNORECASE,
+        ):
             code = canonical.get(match.group(0).lower())
             if code:
                 found.append(code)
@@ -340,6 +388,11 @@ def _species_mentions_with_roles(query: str) -> Dict[str, List[str]]:
         ),
         ("comparativeSpecies", r"比较物种|比较|对比|comparative|comparison|outgroup|外群"),
     ]
+    role_bound_marker_patterns = {
+        "targetSpecies": r"目标物种|目标基因组|target\s+(?:species|genome)",
+        "referenceSpecies": r"参考物种|参考基因组|reference\s+(?:species|genome)",
+        "comparativeSpecies": r"比较物种|比较基因组|comparative\s+(?:species|genome)|outgroup",
+    }
     roles = {"targetSpecies": [], "referenceSpecies": [], "comparativeSpecies": [], "unassigned": []}
     # A role may follow the species, as in "Ft是目标物种". Resolve
     # comma-delimited clauses first so a path fragment such as ref_data cannot
@@ -349,6 +402,15 @@ def _species_mentions_with_roles(query: str) -> Dict[str, List[str]]:
         for role, pattern in marker_patterns:
             if re.search(pattern, clause_lower):
                 roles[role].extend(collect_species(clause))
+                role_bound_marker = role_bound_marker_patterns[role]
+                for role_bound_pattern in (
+                    rf"(?i:(?:{role_bound_marker})\s*(?:为|是|作为|is|[:：=])?\s*)({_ROLE_BOUND_SPECIES_CODE})",
+                    rf"({_ROLE_BOUND_SPECIES_CODE})(?i:\s*(?:是|为|作为|is)?\s*(?:{role_bound_marker}))",
+                ):
+                    roles[role].extend(
+                        match.group(1)
+                        for match in re.finditer(role_bound_pattern, clause)
+                    )
     if any(roles[role] for role in ("targetSpecies", "referenceSpecies", "comparativeSpecies")):
         assigned = set(roles["targetSpecies"] + roles["referenceSpecies"] + roles["comparativeSpecies"])
         roles["unassigned"] = [sp for sp in collect_species(text) if sp not in assigned]
@@ -368,8 +430,7 @@ def _species_mentions_with_roles(query: str) -> Dict[str, List[str]]:
         return {key: _unique_keep_order(value) for key, value in roles.items()}
 
     mentions: List[tuple[int, str]] = []
-    code_pattern = r"(?<![A-Za-z0-9])(?:Ft|At|Os|Ls|Zm|Sl|Lc|Dl|Ppy|Py|Cp|Vv|Lg)(?![A-Za-z0-9])"
-    for match in re.finditer(code_pattern, text, flags=re.IGNORECASE):
+    for match in re.finditer(_SPECIES_CODE_PATTERN, text, flags=re.IGNORECASE):
         code = canonical.get(match.group(0).lower())
         if code:
             mentions.append((match.start(), code))
@@ -646,6 +707,10 @@ def parse_intent_from_query(query: str) -> Dict:
     query_lower = query.lower()
     query_without_space = re.sub(r"\s+", "", query_lower)
 
+    report_intent = resolve_gene_family_report_intent(query)
+    if report_intent:
+        return report_intent
+
     if any(token.replace(" ", "") in query_without_space for token in (
         "wgs全流程", "vcf全流程", "wgs full pipeline", "vcf full pipeline"
     )):
@@ -732,6 +797,10 @@ def parse_intent_from_query(query: str) -> Dict:
         "clarify_type": result.get("clarify_type"),
         "message": result.get("message")
     }
+    if parsed["primary_stage"] == "rnaseq_report":
+        parsed["params"] = {
+            "rnaseqReportAction": "open" if "打开" in query_lower else "generate"
+        }
     if ncbi_params:
         parsed["params"] = ncbi_params
         parsed["prefix"] = ncbi_params.get("prefix", "")
@@ -781,6 +850,43 @@ def parse_intent_from_query(query: str) -> Dict:
     return parsed
 
 
+def deterministic_agent_chat_intent(
+    query: str,
+    *,
+    rnaseq_context: bool = False,
+) -> Optional[Dict]:
+    """Return explicit chat intents that must not depend on the Router model."""
+    intent = parse_intent_from_query(query)
+    if intent.get("stages") == ["rnaseq_report"]:
+        return intent
+    query_lower = str(query or "").lower()
+    query_without_space = re.sub(r"\s+", "", query_lower)
+    is_generic_report_action = (
+        rnaseq_context
+        and "结果报告" in query_without_space
+        and any(action in query_without_space for action in ("生成", "打开"))
+    )
+    if is_generic_report_action:
+        return {
+            "gene_family": "",
+            "species": "",
+            "stages": ["rnaseq_report"],
+            "primary_stage": "rnaseq_report",
+            "task_type": "report",
+            "task_kind": "pipeline",
+            "confidence": 0.99,
+            "need_clarify": False,
+            "clarify_type": None,
+            "message": None,
+            "params": {
+                "rnaseqReportAction": (
+                    "open" if "打开" in query_without_space else "generate"
+                )
+            },
+        }
+    return None
+
+
 def parse_msa_region_plot_request(query: str) -> Dict:
     """Extract parameters for drawing a selected MSA alignment region."""
     text = str(query or "").strip()
@@ -1367,6 +1473,12 @@ def _extract_stages(query: str, query_lower: str) -> List[str]:
         stages.append("gwas_fmt")
     matched_keywords = []  # 记录匹配到的关键词长度，用于优先级判断
     has_rnaseq_marker = any(marker in query_lower for marker in RNASEQ_MARKERS)
+    requests_result_report = any(
+        keyword in query_lower
+        for keyword in ["生成结果报告", "打开结果报告", "结果报告", "生成报告", "打开报告"]
+    )
+    if has_rnaseq_marker and requests_result_report:
+        return ["rnaseq_report"]
     extraction_context = any(kw in query_lower for kw in ["提取", "extract", "gffread", "筛选", "subseq"])
     explicit_gene_family_context = any(kw in query_lower for kw in ["基因家族", "家族表达", "表达量", "热图", "heatmap"])
     curated_member_context = (

gfa_core/mas/workflow_models.py

@@ -2611,6 +2611,50 @@ class WorkflowStateSnapshot(_WorkflowModel):
         return self
 
 
+class CurrentRecommendation(_WorkflowModel):
+    recommendation_id: str
+    job_id: str
+    workflow_id: Literal[
+        "data_preparation",
+        "gene_family",
+        "rnaseq_whole_genome",
+        "rnaseq_family_member",
+    ]
+    transition_id: str
+    from_state_id: str
+    to_state_id: str
+    target_stage: str
+    authorized_stages: tuple[str, ...] = Field(min_length=1)
+    evidence_digest: str
+    family: str | None = None
+    scope_id: str | None = None
+
+    @model_validator(mode="after")
+    def validate_recommendation(self) -> "CurrentRecommendation":
+        for field_name in (
+            "recommendation_id",
+            "job_id",
+            "transition_id",
+            "from_state_id",
+            "to_state_id",
+            "target_stage",
+        ):
+            _require_nonblank(getattr(self, field_name), field_name)
+        if len(self.authorized_stages) != len(set(self.authorized_stages)):
+            raise ValueError("authorized_stages must contain unique stages")
+        if any(not stage.strip() for stage in self.authorized_stages):
+            raise ValueError("authorized_stages cannot contain blank stages")
+        if self.target_stage not in self.authorized_stages:
+            raise ValueError("target_stage must be authorized")
+        if not _SHA256_RE.fullmatch(self.evidence_digest):
+            raise ValueError("evidence_digest must be a sha256 hex digest")
+        for field_name in ("family", "scope_id"):
+            value = getattr(self, field_name)
+            if value is not None:
+                _require_nonblank(value, field_name)
+        return self
+
+
 class GoalLedger(_WorkflowModel):
     goal_ledger_version: Literal["goal_ledger_v1"] = "goal_ledger_v1"
     conversation_id: str

gfa_core/schemas/stages/gene_family_report_contract_v1.md

@@ -0,0 +1,82 @@
+# Gene Family Report Stage Contract v1.0
+
+## 阶段信息
+
+- **阶段名称**: `gene_family_report`
+- **脚本**: `apps/cli/run_gene_family_report.py`
+- **职责**: 按用户确认的计划阶段，生成或打开可持久化的基因家族分析结果报告。
+
+## 输入契约
+
+| 字段 | 类型 | 必需 | 描述 |
+|---|---|---|---|
+| workspace | directory | 是 | 当前 job 工作目录 |
+| prefix | string | 是 | 结果前缀 |
+| expected_stages | non-empty stage list | 生成时是 | 用户确认的计划阶段，保持原顺序；不得从目录中已有 Manifest 反推计划 |
+| stage_statuses | JSON object | 是 | 每个计划阶段的终态或当前状态 |
+| job_id | string | 否 | job 身份 |
+| scope_id / analysis_route / member_set_id | string | 否 | 严格证据身份 |
+| rnaseq_scope_id | string | 否 | 可选 RNA-seq 证据身份 |
+| action | enum | 否 | `generate` 生成；`open` 有现成报告时复用，否则生成 |
+
+阶段不启动任何上游生物分析。至少一个同身份、同计划范围的成功 Manifest 才能生成报告。
+
+`action=generate`，以及 `action=open` 但不存在可复用报告时，`expected_stages` 必须为非空；否则明确拒绝生成。`action=open` 且 Markdown 与报告 Manifest 已存在时可以直接复用，不要求再次提供计划。
+
+## 输出契约
+
+| artifact key | 类型 | 文件 |
+|---|---|---|
+| gene_family_report | markdown | `*.gene_family_report.md` |
+| gene_family_report_manifest | manifest | `*.gene_family_report.manifest.json` |
+
+Manifest 记录 `final|partial` 报告状态、`api|fallback_script` 生成模式和当前分析身份。`included_stages` 只包含模板支持且成功的计划阶段；成功但不在模板 13 个阶段中的计划项写入 `unsupported_stages`，在公共完成表标记“不支持”，不生成推断章节，并使报告状态保持 `partial`。API 不可用或编排越界时，`model.reason` 固定为 `api_unavailable_or_boundary_violation`。
+
+## 确定性数据接口
+
+`build_report_data(final_dir, prefix, planned, statuses, manifests) -> ReportData` 是模板渲染和 API 编排共用的唯一结果模型。每节把脚本确定的内容分为 `facts`、`observations`、`interpretations`、`limitations`、`recommendations`，并汇总为带稳定 `sentence_id`、原文、类型和证据编号的不可变句子。
+
+指标优先读取同身份 Manifest 的嵌套 `metrics`。仅当模板所需指标缺失时，集中提取器才读取该指标声明的命名 artifact；不得从文件名、其他阶段或相似字段猜值。真缺失使用 `MissingMetric(label, checked_sources)`，渲染为“源结果未提供此指标（已检查：…）”。声明的数据源存在但解析失败时必须报错，不得静默写成“未解析”、`N/A` 或 `None`。
+
+当前真实 schema 映射包括：Identify 的 `member_count` / `identified_members_count` / `final_members`（按此前顺序取首个已提供值），Phylogeny 的 `sequences_count` 与 `msa_length`，PEP 的 `protein_count`、`tasks_success` 与 `tasks_total`。Interspecies Synteny 在没有直接语义指标时只读取 Manifest 声明的 `gene_to_species` 与 `*_collinearity` 命名 artifact：以 `## Alignment` 计共线性区块，并按 TSV 的 `group=target|ref` 名单统计目标—参考家族直接配对。
+
+## API 语言编排接口
+
+API 输入字段仅限：
+
+```json
+{
+  "sections": [
+    {
+      "section_id": "stage.identify",
+      "sentences": [{"sentence_id": "stage.identify.status", "text": "不可改写原句"}]
+    }
+  ],
+  "allowed_connectors": ["also", "however", "therefore"]
+}
+```
+
+API 输出字段仅限：
+
+```json
+{
+  "sections": [
+    {
+      "section_id": "stage.identify",
+      "paragraphs": [
+        {"sentence_ids": ["stage.identify.status"], "connector_ids": []}
+      ]
+    }
+  ]
+}
+```
+
+输出必须保持完整章节集合；每个不可变句子必须在原章节中恰好出现一次；连接语必须来自白名单，数量必须与段内句间位置一致。未知、遗漏或重复 `sentence_id`，未知连接语，额外字段，自由文本字段，章节遗漏、重复或改变，均使整份 API 编排作废并回退 `fallback_script`，不能局部采用。
+
+## 报告生成原则
+
+1. Manifest、结果表格和结果文件是报告事实的唯一来源。成员数、统计值、比较结果、证据等级、限制、可下结论范围和后续建议均由确定性脚本提取或生成。
+2. 模型 API 默认启用，但只负责排列脚本已生成的 `sentence_id`、组织段落并选择受控 `connector_id`；API 不直接返回报告文本。不得补充、推断或修改任何数值、结果判断、生物学解释、因果关系、候选基因、证据等级、限制或建议。
+3. API 输出必须通过上述严格结构与集合校验；任一边界错误都使整份 API 编排作废，并自动使用同一 `ReportData` 的默认顺序。
+4. API 不可用、超时或校验失败时，不影响报告事实和章节完整性；报告明确标注 `fallback_script`，但仍使用同一模板和同一份确定性结果数据。
+5. 只纳入用户计划中的阶段。源结果确实没有提供某项指标时，明确写“源结果未提供此指标”并指出所检查的来源；不得用“未解析”、`N/A` 或模型猜测代替可提取的数据。

gfa_core/templates/gene_family_report_v1.md

@@ -0,0 +1,151 @@
+<!-- GFA_SECTION common.front -->
+# {{PREFIX}} 基因家族分析结果解读报告
+
+- 任务：{{JOB_ID}}
+- 分析前缀：{{PREFIX}}
+- 生成时间：{{GENERATED_AT}}
+
+
+## 一、报告首页
+
+### 1.1 项目信息
+
+本报告只覆盖用户计划中的成功阶段；不同阶段的目标集合在各节单独标明。
+
+### 1.2 研究目的
+
+本报告用于确认家族成员，比较目标成员及必要参考成员的关系、结构、调控序列与蛋白预测特征，并界定结果能够支持的结论范围。
+
+### 1.3 一页式结论摘要
+
+{{SECTION_CONTENT}}
+
+### 1.4 结论可信度说明
+
+成员数和命名表中的统计可直接复核；树、启动子、TFBS、共线性和蛋白定位只提供计算线索，不等同于功能或因果证据。
+
+<!-- GFA_SECTION common.route -->
+## 二、分析路线与阅读顺序
+
+{{SECTION_CONTENT}}
+
+### 2.1 统一解读顺序
+
+各阶段依次写明分析对象、源结果中的直接观察、谨慎解释、限制，并仅在有助复核时列出少量相对路径。
+
+### 2.2 语言边界
+
+预测关系统一表述为线索；没有逐成员表或独立证据时，不指定候选基因，不声称功能改变或调控关系已经证实。
+
+<!-- GFA_SECTION common.completeness -->
+## 三、结果总览与完整性检查
+
+### 3.1 阶段完成情况
+
+{{SECTION_CONTENT}}
+
+### 3.2 结果完整性概述
+
+报告按计划逐阶段选择成功 Manifest；当各阶段使用的成员集合不同，正文明确区分目标成员、目标加参考成员和全基因组背景。
+
+<!-- GFA_SECTION stage.identify -->
+## 四、基因家族鉴定（Identify）
+
+### 4.3 最终成员数量与命名
+
+{{SECTION_CONTENT}}
+
+<!-- GFA_SECTION stage.phylogeny -->
+## 五、亲缘关系分析（Phylogeny）
+
+### 5.5 亲缘关系阶段小结
+
+{{SECTION_CONTENT}}
+
+<!-- GFA_SECTION stage.reference_classification -->
+## 六、参考物种辅助的小组分类（Reference Classification）
+
+### 6.2 分类可信度表述
+
+{{SECTION_CONTENT}}
+
+<!-- GFA_SECTION stage.motif -->
+## 七、短保守片段分析（Motif）
+
+### 7.2 Motif 数量与出现频率
+
+{{SECTION_CONTENT}}
+
+<!-- GFA_SECTION stage.domain -->
+## 八、主要功能区域分析（Domain）
+
+### 8.1 Domain 整体统计
+
+{{SECTION_CONTENT}}
+
+<!-- GFA_SECTION stage.structure -->
+## 九、基因内部结构分析（Structure）
+
+### 9.1 结构覆盖与异常结构
+
+{{SECTION_CONTENT}}
+
+<!-- GFA_SECTION stage.combo -->
+## 十、综合组合图（Combo）
+
+### 10.1 树 + Motif + Domain + Structure 组合图
+
+{{SECTION_CONTENT}}
+
+<!-- GFA_SECTION stage.promoter -->
+## 十一、启动子元件分析（Promoter）
+
+### 11.1 元件总量、类别与高级统计
+
+{{SECTION_CONTENT}}
+
+<!-- GFA_SECTION stage.tfbs -->
+## 十二、转录因子结合位点预测（TFBS）
+
+### 12.1 高频结合片段与统计边界
+
+{{SECTION_CONTENT}}
+
+<!-- GFA_SECTION stage.synteny -->
+## 十三、种内共线性分析（Synteny）
+
+### 13.1 全基因组共线性背景与家族证据缺口
+
+{{SECTION_CONTENT}}
+
+<!-- GFA_SECTION stage.karyotype -->
+## 十四、染色体分布（Karyotype）
+
+### 14.1 染色体位置概述
+
+{{SECTION_CONTENT}}
+
+<!-- GFA_SECTION stage.interspecies_synteny -->
+## 十五、种间共线性与跨物种关系（Interspecies Synteny）
+
+### 15.6 Interspecies Synteny 阶段小结
+
+{{SECTION_CONTENT}}
+
+<!-- GFA_SECTION stage.pep -->
+## 十六、蛋白基本特征分析（PEP）
+
+### 16.1 蛋白理化性质分布图
+
+逐蛋白理化性质保留在命名结果表中；源 Manifest 没有汇总长度、分子量或等电点范围时，正文不从图片反推数值。
+
+### 16.4 亚细胞定位与阶段小结
+
+{{SECTION_CONTENT}}
+
+<!-- GFA_SECTION common.integrated -->
+## 十七、多证据综合分析
+
+### 17.5 综合结论
+
+{{SECTION_CONTENT}}

manage_plantai.sh

@@ -2,15 +2,19 @@
 # PlantAI-dev 9000 服务管理脚本
 # 用法: bash manage_plantai.sh {start|stop|restart|status} [--port PORT]
 # 默认端口: 9000
-#
-# 项目: /home/xzq/5090/teamWorkspace/zh/PlantAI-dev
-# 端口: 9000 (web), Redis DB 14 (queue prefix: plantai_dev_9002)
 
 set -euo pipefail
 
+SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
+if [ -f "$SCRIPT_DIR/.env" ]; then
+    set -a
+    . "$SCRIPT_DIR/.env"
+    set +a
+fi
+
 # 解析命令行参数：第一个参数是命令，后面可选 --port / -p
 CMD=""
-WEB_PORT="9000"
+WEB_PORT="${PLANTAI_WEB_PORT:-9000}"
 for arg in "$@"; do
     case "$arg" in
         start|stop|restart|status)
@@ -33,12 +37,15 @@ done
 CMD="${CMD:-status}"
 
 # Conda 环境：启动脚本内部也依赖这个环境
-CONDA_ENV="${CONDA_ENV:-/home/xzq/miniconda3/envs/test4py311}"
+CONDA_BASE="${BBAV5_CONDA_BASE:-/home/xia-lab/miniconda3}"
+DEFAULT_ENV="${BBAV5_DEFAULT_ENV:-test4py311}"
+CONDA_ENV="${CONDA_ENV:-$CONDA_BASE/envs/$DEFAULT_ENV}"
 export PATH="$CONDA_ENV/bin:$PATH"
 
 # 项目配置（与 .env 和启动脚本保持一致）
-PROJECT_DIR="/home/xzq/5090/teamWorkspace/zh/PlantAI-dev"
-QUEUE_PREFIX="plantai_dev_9002"
+PROJECT_DIR="$SCRIPT_DIR"
+QUEUE_PREFIX="${BBAV5_QUEUE_PREFIX:-plantai_4090_9000}"
+REDIS_URL="${BBAV5_REDIS_URL:-redis://localhost:6379/10}"
 WEB_LOG="/tmp/plantai_${WEB_PORT}.log"
 CELERY_LOG="/tmp/celery_${WEB_PORT}.log"
 
@@ -46,28 +53,49 @@ red()  { echo -e "\033[31m$*\033[0m"; }
 green(){ echo -e "\033[32m$*\033[0m"; }
 cyan() { echo -e "\033[36m$*\033[0m"; }
 
+web_pid_for_port() {
+    ss -tlnp 2>/dev/null \
+        | grep ":$WEB_PORT " \
+        | grep -oP 'pid=\K\d+' \
+        | head -n 1 \
+        || true
+}
+
+web_process_cwd() {
+    readlink -f "/proc/$1/cwd" 2>/dev/null || true
+}
+
+web_pid_belongs_to_project() {
+    [ "$(web_process_cwd "$1")" = "$PROJECT_DIR" ]
+}
+
 status() {
     echo ""
     cyan "=== PlantAI-dev ${WEB_PORT} 服务状态 ==="
     echo ""
 
     # Web 服务
-    WEB_PID=$(ss -tlnp 2>/dev/null | grep ":$WEB_PORT " | grep -oP 'pid=\K\d+' || true)
+    WEB_PID="$(web_pid_for_port)"
     if [ -n "${WEB_PID:-}" ]; then
-        green "[✓] Web  服务运行中  (端口 $WEB_PORT, PID $WEB_PID)"
+        WEB_CWD="$(web_process_cwd "$WEB_PID")"
+        if web_pid_belongs_to_project "$WEB_PID"; then
+            green "[✓] Web  服务运行中  (端口 $WEB_PORT, PID $WEB_PID)"
+        else
+            red "[!] 端口 $WEB_PORT 被其他项目占用  (PID $WEB_PID, cwd ${WEB_CWD:-未知})"
+        fi
     else
         red "[✗] Web  服务未启动  (端口 $WEB_PORT)"
     fi
 
     # Redis
-    if redis-cli ping > /dev/null 2>&1; then
-        green "[✓] Redis 运行中  (localhost:6379)"
+    if redis-cli -u "$REDIS_URL" ping > /dev/null 2>&1; then
+        green "[✓] Redis 运行中  ($REDIS_URL)"
     else
         red "[✗] Redis 未启动"
     fi
 
     # Celery workers
-    CELERY_COUNT=$(pgrep -fc "$QUEUE_PREFIX" || echo 0)
+    CELERY_COUNT="$(pgrep -fc "$QUEUE_PREFIX" || true)"
     if [ "$CELERY_COUNT" -gt 0 ]; then
         green "[✓] Celery workers 运行中  (${CELERY_COUNT} 进程, 队列前缀: $QUEUE_PREFIX)"
         echo "    Worker 列表:"
@@ -86,13 +114,18 @@ stop() {
     echo ""
 
     # 停止 Web 服务
-    WEB_PID=$(ss -tlnp 2>/dev/null | grep ":$WEB_PORT " | grep -oP 'pid=\K\d+' || true)
+    WEB_PID="$(web_pid_for_port)"
     if [ -n "${WEB_PID:-}" ]; then
+        WEB_CWD="$(web_process_cwd "$WEB_PID")"
+        if ! web_pid_belongs_to_project "$WEB_PID"; then
+            red "[!] 拒绝停止：端口 $WEB_PORT 属于其他项目  (PID $WEB_PID, cwd ${WEB_CWD:-未知})"
+            return 1
+        fi
         echo "正在停止 Web 服务 (PID $WEB_PID)..."
         kill "$WEB_PID" 2>/dev/null || true
         sleep 1
-        # 如果还没死，强制 kill
-        if ss -tlnp 2>/dev/null | grep -q ":$WEB_PORT"; then
+        # 仅强制停止仍属于本项目的原进程，避免 PID 复用造成误杀。
+        if kill -0 "$WEB_PID" 2>/dev/null && web_pid_belongs_to_project "$WEB_PID"; then
             kill -9 "$WEB_PID" 2>/dev/null || true
         fi
         green "[✓] Web 服务已停止"
@@ -125,12 +158,18 @@ start() {
     echo ""
 
     # 检查是否已在运行
-    if ss -tlnp 2>/dev/null | grep -q ":$WEB_PORT"; then
-        red "[!] 端口 $WEB_PORT 已被占用，请先停止服务"
+    WEB_PID="$(web_pid_for_port)"
+    if [ -n "${WEB_PID:-}" ]; then
+        WEB_CWD="$(web_process_cwd "$WEB_PID")"
+        if web_pid_belongs_to_project "$WEB_PID"; then
+            red "[!] 本项目 Web 已在端口 $WEB_PORT 运行  (PID $WEB_PID)"
+        else
+            red "[!] 端口 $WEB_PORT 被其他项目占用  (PID $WEB_PID, cwd ${WEB_CWD:-未知})"
+        fi
         exit 1
     fi
 
-    if ! redis-cli ping > /dev/null 2>&1; then
+    if ! redis-cli -u "$REDIS_URL" ping > /dev/null 2>&1; then
         red "[!] Redis 未运行，请先启动 Redis"
         exit 1
     fi
@@ -141,7 +180,7 @@ start() {
     echo "正在启动 Celery workers..."
     nohup bash start_celery_9000.sh > "$CELERY_LOG" 2>&1 &
     sleep 3
-    CELERY_COUNT=$(pgrep -fc "$QUEUE_PREFIX" || echo 0)
+    CELERY_COUNT="$(pgrep -fc "$QUEUE_PREFIX" || true)"
     if [ "$CELERY_COUNT" -gt 0 ]; then
         green "  [✓] Celery workers 已启动 (${CELERY_COUNT} 进程)"
     else

start_celery_9000.sh

@@ -26,13 +26,13 @@ if [ -n "$EXTERNAL_BBAV5_WORKER_NAME_PREFIX" ]; then
     BBAV5_WORKER_NAME_PREFIX="$EXTERNAL_BBAV5_WORKER_NAME_PREFIX"
 fi
 if [ -z "${BBAV5_QUEUE_PREFIX-}" ]; then
-    BBAV5_QUEUE_PREFIX="plantai_dev_9000"
+    BBAV5_QUEUE_PREFIX="plantai_4090_9000"
 fi
 if [ -z "${BBAV5_REDIS_URL-}" ]; then
     BBAV5_REDIS_URL="redis://localhost:6379/10"
 fi
 if [ -z "${BBAV5_CELERY_LOG_DIR-}" ]; then
-    BBAV5_CELERY_LOG_DIR="logs/celery_plantai_dev_9000"
+    BBAV5_CELERY_LOG_DIR="logs/celery_plantai_4090_9000"
 fi
 if [ -z "$EXTERNAL_BBAV5_WORKER_NAME_PREFIX" ]; then
     unset BBAV5_WORKER_NAME_PREFIX
@@ -51,11 +51,11 @@ export BBAV5_PROJECT_ENV_FILE="${BBAV5_PROJECT_ENV_FILE:-$PROJECT_ROOT/.env}"
 export BBAV5_LIGHTWEIGHT_EXTENSION_STORE="${BBAV5_LIGHTWEIGHT_EXTENSION_STORE:-$PROJECT_ROOT/.runtime/lightweight-extension/store}"
 export BBAV5_LIGHTWEIGHT_INPUT="${BBAV5_LIGHTWEIGHT_INPUT:-$PROJECT_ROOT/input}"
 export BBAV5_LIGHTWEIGHT_WORKSPACE="${BBAV5_LIGHTWEIGHT_WORKSPACE:-$PROJECT_ROOT/workspace}"
-export BBAV5_QUEUE_PREFIX="${BBAV5_QUEUE_PREFIX:-plantai_dev_9000}"
+export BBAV5_QUEUE_PREFIX="${BBAV5_QUEUE_PREFIX:-plantai_4090_9000}"
 export BBAV5_REDIS_URL="${BBAV5_REDIS_URL:-redis://localhost:6379/10}"
 WORKER_NAME_PREFIX="${BBAV5_WORKER_NAME_PREFIX:-${BBAV5_QUEUE_PREFIX//[^A-Za-z0-9_]/_}}"
 
-LOG_DIR="${BBAV5_CELERY_LOG_DIR:-logs/celery_plantai_dev_9000}"
+LOG_DIR="${BBAV5_CELERY_LOG_DIR:-logs/celery_plantai_4090_9000}"
 mkdir -p "$LOG_DIR"
 
 CELERY="$CONDA_ENV/bin/celery"

start_plantai_9000.sh

@@ -24,6 +24,12 @@ done
 SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)"
 cd "$SCRIPT_DIR"
 
+if [ -f "$SCRIPT_DIR/.env" ]; then
+    set -a
+    . "$SCRIPT_DIR/.env"
+    set +a
+fi
+
 CONDA_BASE="${BBAV5_CONDA_BASE:-/home/xia-lab/miniconda3}"
 DEFAULT_ENV="${BBAV5_DEFAULT_ENV:-test4py311}"
 CONDA_ENV="$CONDA_BASE/envs/$DEFAULT_ENV"
@@ -33,23 +39,18 @@ export PATH="$CONDA_ENV/bin:$PATH"
 # the caller's shell; GWAS/other sandboxes have dedicated restart scripts.
 export BBAV5_PROJECT_ROOT="$SCRIPT_DIR"
 export PLANTAI_PROJECT_ROOT="$SCRIPT_DIR"
-if [ -f "$SCRIPT_DIR/.env" ]; then
-    set -a
-    . "$SCRIPT_DIR/.env"
-    set +a
-fi
 export BBAV5_PROJECT_ENV_FILE="${BBAV5_PROJECT_ENV_FILE:-$SCRIPT_DIR/.env}"
 export BBAV5_LIGHTWEIGHT_EXTENSION_STORE="${BBAV5_LIGHTWEIGHT_EXTENSION_STORE:-$SCRIPT_DIR/.runtime/lightweight-extension/store}"
 export BBAV5_LIGHTWEIGHT_INPUT="${BBAV5_LIGHTWEIGHT_INPUT:-$SCRIPT_DIR/input}"
 export BBAV5_LIGHTWEIGHT_WORKSPACE="${BBAV5_LIGHTWEIGHT_WORKSPACE:-$SCRIPT_DIR/workspace}"
-export BBAV5_QUEUE_PREFIX="${BBAV5_QUEUE_PREFIX:-plantai_dev_9000}"
+export BBAV5_QUEUE_PREFIX="${BBAV5_QUEUE_PREFIX:-plantai_4090_9000}"
 export BBAV5_REDIS_URL="${BBAV5_REDIS_URL:-redis://localhost:6379/10}"
 export PLANTAI_WEB_PORT="$PORT"
 export PLANTAI_RESTART_SCRIPT="$SCRIPT_DIR/restart_plantai_9000.sh"
-export GENE_FAMILY_MAS_API="http://127.0.0.1:7926"
-export DUAL_KB_QA_BASE_URL="http://127.0.0.1:7926"
-export RAG_API_URL="http://127.0.0.1:7926/api/v1/qa-v1-1-73/chat"
-export DUAL_KB_QA_ROUTE="qa-v1-1-73"
+export GENE_FAMILY_MAS_API="${GENE_FAMILY_MAS_API:-${DUAL_KB_QA_BASE_URL:-http://127.0.0.1:7920}}"
+export DUAL_KB_QA_BASE_URL="${DUAL_KB_QA_BASE_URL:-$GENE_FAMILY_MAS_API}"
+export RAG_API_URL="${RAG_API_URL:-$GENE_FAMILY_MAS_API}"
+export DUAL_KB_QA_ROUTE="${DUAL_KB_QA_ROUTE:-qa-v1-1-73}"
 export PLANTAI_AGENT_STATE_DB="$SCRIPT_DIR/.runtime/agent-state/agent-state.sqlite3"
 
 PORT="$PLANTAI_WEB_PORT"

tasks.py

@@ -411,7 +411,7 @@ def validate_single_stage_artifacts(job_id: str, prefix: str, stage: str, result
         stage_status = get_stage_status_for_identity(
             job_id=job_id,
             identity=identity,
-            strict_identity=bool(
+            strict_identity=actual_stage != "gene_family_report" and bool(
                 identity.scope_id
                 or identity.analysis_route
                 or identity.member_set_id

tests/e2e/verify_gene_family_report_journey.py

@@ -0,0 +1,226 @@
+#!/usr/bin/env python3
+"""Small real-data journey for the gene-family report module."""
+
+from __future__ import annotations
+
+import argparse
+import json
+from pathlib import Path
+import re
+import shutil
+import sys
+
+
+PROJECT_ROOT = Path(__file__).resolve().parents[2]
+if str(PROJECT_ROOT) not in sys.path:
+    sys.path.insert(0, str(PROJECT_ROOT))
+
+from apps.cli.run_gene_family_report import main as report_cli_main
+from gfa_core.mas.utils.intent_parser import parse_intent_from_query
+
+
+PREFIX = "Real.SPL"
+SCOPE_ID = "scope-real-spl"
+ANALYSIS_ROUTE = "post_subfamily_confirm"
+MEMBER_SET_ID = "members-real-spl"
+
+
+def _sequence_count(path: Path) -> int:
+    count = sum(line.startswith(">") for line in path.read_text(encoding="utf-8").splitlines())
+    if not count:
+        raise AssertionError("representative FASTA is empty")
+    return count
+
+
+def _write_identify_manifest(workspace: Path, fasta: Path, count: int) -> None:
+    final_dir = workspace / "final"
+    final_dir.mkdir(parents=True, exist_ok=True)
+    manifest = {
+        "manifest_version": "1.0",
+        "stage": "identify",
+        "status": "success",
+        "prefix": PREFIX,
+        "scope_id": SCOPE_ID,
+        "analysis_route": ANALYSIS_ROUTE,
+        "member_set_id": MEMBER_SET_ID,
+        "parameters": {
+            "scope_id": SCOPE_ID,
+            "analysis_route": ANALYSIS_ROUTE,
+            "member_set_id": MEMBER_SET_ID,
+        },
+        "metrics": {"member_count": count, "final_members": count, "total_proteins": count},
+        "artifacts": {
+            "final_members": {"path": str(fasta), "full_path": str(fasta), "type": "protein_fasta"}
+        },
+    }
+    (final_dir / f"{PREFIX}.identify.manifest.json").write_text(
+        json.dumps(manifest, ensure_ascii=False, indent=2) + "\n",
+        encoding="utf-8",
+    )
+
+
+def _run(workspace: Path, action: str):
+    return report_cli_main(
+        [
+            "--workspace", str(workspace),
+            "--prefix", PREFIX,
+            "--job-id", "representative-gene-family-report",
+            "--expected-stages", "identify",
+            "--stage-statuses-json", '{"identify":"success"}',
+            "--scope-id", SCOPE_ID,
+            "--analysis-route", ANALYSIS_ROUTE,
+            "--member-set-id", MEMBER_SET_ID,
+            "--action", action,
+            "--no-model",
+        ]
+    )
+
+
+def _stage_section(report: str, stage: str) -> str:
+    heading = re.search(rf"(?m)^## (?!#).*?\(`{re.escape(stage)}`\)", report)
+    if heading is None:
+        raise AssertionError(f"missing report chapter for {stage}")
+    next_heading = re.search(r"(?m)^## (?!#)", report[heading.end():])
+    end = heading.end() + next_heading.start() if next_heading else len(report)
+    return report[heading.start():end]
+
+
+def _subsection(report: str, number: str) -> str:
+    heading = re.search(rf"(?m)^### {re.escape(number)}(?:\s|$)", report)
+    if heading is None:
+        raise AssertionError(f"missing report subsection {number}")
+    next_heading = re.search(r"(?m)^#{2,3} ", report[heading.end():])
+    end = heading.end() + next_heading.start() if next_heading else len(report)
+    return report[heading.start():end]
+
+
+def _verify_real_job(workspace: Path, prefix: str) -> None:
+    final_dir = workspace / "final"
+    report_path = final_dir / f"{prefix}.gene_family_report.md"
+    manifest_path = final_dir / f"{prefix}.gene_family_report.manifest.json"
+    report = report_path.read_text(encoding="utf-8")
+    manifest = json.loads(manifest_path.read_text(encoding="utf-8"))
+    expected_stages = [
+        "identify",
+        "phylogeny",
+        "reference_classification",
+        "motif",
+        "domain",
+        "structure",
+        "combo",
+        "promoter",
+        "tfbs",
+        "synteny",
+        "karyotype",
+        "interspecies_synteny",
+        "pep",
+    ]
+
+    assert report.startswith(f"# {prefix} 基因家族分析结果解读报告\n")
+    for placeholder in ("【填写】", "未解析", "N/A", "None"):
+        assert placeholder not in report
+    assert "/home/xzq/" not in report
+    assert not re.search(r"(?m)^- `[a-zA-Z0-9_.-]+`:\s*", report)
+    assert manifest["parameters"]["expected_stages"] == expected_stages
+    assert manifest["metrics"]["included_stages"] == expected_stages
+    assert manifest["metrics"]["generation_mode"] == "fallback_script"
+    assert sorted(path.name for path in final_dir.glob(f"{prefix}.gene_family_report*")) == [
+        f"{prefix}.gene_family_report.manifest.json",
+        f"{prefix}.gene_family_report.md",
+    ]
+    for stage in expected_stages:
+        section = _stage_section(report, stage)
+        for chunk in re.split(r"(?m)(?=^### )", section)[1:]:
+            assert "\n".join(chunk.splitlines()[1:]).strip(), chunk.splitlines()[0]
+        assert "分析对象：" in section
+        assert "限制：" in section
+    assert report.count("本节解释仅限上述可追溯结果，不据此断言真实功能或因果关系。") <= 1
+    summary = _subsection(report, "1.3")
+    assert 6 <= summary.count("。") <= 10
+    assert re.search(r"目标物种.*24.*家族成员", summary)
+    assert re.search(r"形成\s*24\s*个\s*SPL家族", _subsection(report, "4.3"))
+    assert re.search(
+        r"使用\s*40\s*条序列.*比对长度为\s*1449", _stage_section(report, "phylogeny")
+    )
+    motif = _stage_section(report, "motif")
+    assert re.search(r"识别\s*10\s*个\s*motif", motif)
+    assert re.search(r"总位点数为\s*105", motif)
+    assert re.search(r"宽度范围为\s*27–138", motif)
+    assert re.search(r"将\s*24\s*个家族成员定位到\s*8\s*条染色体", _stage_section(report, "karyotype"))
+    assert re.search(
+        r"全基因组\s*block\s*521\s*个.*目标-参考家族.*?2\s*对",
+        _stage_section(report, "interspecies_synteny"),
+    )
+    assert re.search(r"覆盖\s*40\s*个蛋白，子任务成功数为\s*3/3", _stage_section(report, "pep"))
+    assert re.search(r"24\s*个目标成员.*8\s*个小组.*高信心\s*14.*中信心\s*7.*低信心\s*3", _subsection(report, "6.2"))
+    assert re.search(r"40/40.*SBP.*40.*Ig_SPL7.*12", _subsection(report, "8.1"))
+    assert re.search(r"24/40.*FtPinG0003446200\.01\.T01.*12454", _subsection(report, "9.1"))
+    assert "Motif、Domain、Structure" in _subsection(report, "10.1")
+    promoter = _subsection(report, "11.1")
+    assert re.search(r"23\s*个目标基因.*3699.*TATA-box.*1460", promoter)
+    assert "CAAT-box—MYB" in promoter and "FtPinG0006337300.01.T01" in promoter
+    tfbs = _subsection(report, "12.1")
+    assert re.search(r"24\s*条目标启动子.*3734.*AHYPO_015621.*153", tfbs)
+    assert re.search(r"背景启动子为\s*0.*不能.*富集", tfbs)
+    synteny = _subsection(report, "13.1")
+    assert re.search(r"全基因组.*316.*3267.*8\s*条染色体", synteny)
+    assert "源结果未提供家族专属复制基因对" in synteny
+    assert re.search(r"Nucleus\s*37.*Cytoplasm\s*2.*Chloroplast\s*1", _subsection(report, "16.4"))
+
+
+def main() -> int:
+    parser = argparse.ArgumentParser()
+    parser.add_argument("--mode", choices=("data", "e2e"))
+    parser.add_argument("--workspace", type=Path)
+    parser.add_argument("--prefix")
+    args = parser.parse_args()
+
+    if args.workspace is not None:
+        if not args.prefix:
+            parser.error("--prefix is required with --workspace")
+        _verify_real_job(args.workspace, args.prefix)
+        print(json.dumps({"status": "passed", "workspace": str(args.workspace)}, ensure_ascii=False))
+        return 0
+    if args.mode is None:
+        parser.error("--mode is required when --workspace is not provided")
+
+    output_dir = PROJECT_ROOT / "test-results/gene-family-report" / args.mode
+    if output_dir.exists():
+        shutil.rmtree(output_dir)
+    workspace = output_dir / "job"
+    fasta = PROJECT_ROOT / "tests/ft_data/SPL_Ath.fasta"
+    count = _sequence_count(fasta)
+    _write_identify_manifest(workspace, fasta, count)
+
+    generated = _run(workspace, "generate")
+    report = generated.report_path.read_text(encoding="utf-8")
+    manifest = json.loads(generated.manifest_path.read_text(encoding="utf-8"))
+    assert generated.report_path.name == f"{PREFIX}.gene_family_report.md"
+    assert generated.manifest_path.name == f"{PREFIX}.gene_family_report.manifest.json"
+    assert "生成模式：候补脚本" in report
+    assert f"最终成员数：{count}" in report
+    assert "## 基因家族鉴定 (`identify`)" in report
+    assert "(`motif`)" not in report
+    assert manifest["scope_id"] == SCOPE_ID
+    assert manifest["analysis_route"] == ANALYSIS_ROUTE
+    assert manifest["member_set_id"] == MEMBER_SET_ID
+    assert manifest["metrics"]["included_stages"] == ["identify"]
+    assert set(manifest["artifacts"]) == {"gene_family_report", "gene_family_report_manifest"}
+
+    reopened = _run(workspace, "open")
+    assert reopened.reused is True
+    intent = parse_intent_from_query("打开基因家族结果报告")
+    assert intent["stages"] == ["gene_family_report"]
+    assert intent["params"]["reportAction"] == "open"
+
+    print(json.dumps({
+        "status": "passed",
+        "mode": args.mode,
+        "sequence_count": count,
+        "report": str(generated.report_path),
+    }, ensure_ascii=False))
+    return 0
+
+
+if __name__ == "__main__":
+    raise SystemExit(main())

tests/unit/test_agent_chat_actions.py

@@ -92,6 +92,7 @@ async def test_agent_chat_maps_router_actions_and_passes_context(
             "history": history,
             "session_id": "session-7",
             "job_id": "job-7",
+            "language": "zh",
         }
     ]
     assert result["success"] is True
@@ -108,6 +109,40 @@ async def test_agent_chat_maps_router_actions_and_passes_context(
         assert result["intent"]["stages"] == []
 
 
+@pytest.mark.asyncio
+async def test_agent_chat_keeps_english_language_separate_from_business_query(
+    monkeypatch,
+    web_main,
+):
+    captured = []
+    _install_router(
+        monkeypatch,
+        {
+            "status": "ok",
+            "action": "reply",
+            "message": "I can help with that.",
+            "intent": None,
+            "sources": [],
+            "tool_trace": [],
+        },
+        captured,
+    )
+    request = web_main.AgentChatRequest(
+        message="执行接下来的分析",
+        language="en",
+        session_id="english-business-query",
+        job_id="job-english-business-query",
+        history=[],
+    )
+
+    result = await web_main.agent_chat(request)
+
+    assert result["intent"]["message"] == "I can help with that."
+    assert request.message == "执行接下来的分析"
+    assert captured[0]["query"] == "执行接下来的分析"
+    assert captured[0]["language"] == "en"
+
+
 @pytest.mark.asyncio
 async def test_agent_chat_persists_router_proposal_without_analysis_intent(
     monkeypatch,
@@ -158,6 +193,68 @@ async def test_agent_chat_persists_router_proposal_without_analysis_intent(
     }
 
 
+@pytest.mark.asyncio
+async def test_agent_chat_exposes_and_persists_current_recommendation(
+    monkeypatch,
+    web_main,
+    tmp_path,
+):
+    repository = SQLiteAgentStateRepository(tmp_path / "agent-state.sqlite3")
+    repository.initialize()
+    store = PersistentConversationStore(repository)
+    monkeypatch.setattr(web_main, "agent_conversation_store", store)
+    recommendation = {
+        "recommendation_id": "rec-build-subfamily-evidence",
+        "job_id": "job-recommendation",
+        "workflow_id": "gene_family",
+        "transition_id": "build_subfamily_evidence",
+        "from_state_id": "family_identified",
+        "to_state_id": "subfamily_evidence_ready",
+        "target_stage": "combo",
+        "authorized_stages": [
+            "domain",
+            "motif",
+            "phylogeny",
+            "structure",
+            "combo",
+        ],
+        "evidence_digest": "a" * 64,
+        "family": "SPL",
+        "scope_id": "SPL.target-Ft.target_only",
+    }
+    _install_router(
+        monkeypatch,
+        {
+            "status": "ok",
+            "action": "show_result",
+            "message": "下一步是构建亚家族证据。",
+            "intent": None,
+            "recommendation": recommendation,
+            "sources": [],
+            "tool_trace": [],
+        },
+        [],
+    )
+
+    result = await web_main.agent_chat(
+        web_main.AgentChatRequest(
+            message="接下来的分析有哪些？",
+            session_id="recommendation-session",
+            job_id="job-recommendation",
+            history=[],
+        )
+    )
+
+    assert result["action"] == "show_result"
+    assert result["recommendation"] == recommendation
+    ledger = repository.get_goal_ledger(result["session_id"])
+    assert ledger is not None
+    assert (
+        ledger.model_dump(mode="json")["user_terms"]["current_recommendation"]
+        == recommendation
+    )
+
+
 @pytest.mark.asyncio
 async def test_agent_chat_exposes_confirmed_request_for_original_planner(
     monkeypatch,

tests/unit/test_authorized_workflow_action_plan.py

@@ -0,0 +1,137 @@
+import asyncio
+
+from apps.web.vue3 import main as api
+
+
+def test_build_subfamily_evidence_plan_keeps_confirmed_stage_boundary():
+    intent = {
+        "workflow_action_id": "build_subfamily_evidence",
+        "stages": ["domain", "motif", "phylogeny", "structure", "combo"],
+        "params": {
+            "workflowActionId": "build_subfamily_evidence",
+            "authorizedStages": [
+                "domain",
+                "motif",
+                "phylogeny",
+                "structure",
+                "combo",
+            ],
+        },
+    }
+
+    stages = api._authorized_workflow_action_stages(intent)
+
+    assert stages == ["phylogeny", "motif", "domain", "structure", "combo"]
+    assert "identify" not in stages
+    assert "reference_classification" not in stages
+
+
+def test_unbound_plan_does_not_invent_an_authorized_stage_boundary():
+    assert api._authorized_workflow_action_stages(
+        {"stages": ["identify", "phylogeny", "reference_classification"]}
+    ) == []
+
+
+def test_agent_plan_honors_authorized_workflow_action_without_keyword_reclassification(
+    monkeypatch,
+):
+    authorized = ["domain", "motif", "phylogeny", "structure", "combo"]
+    intent = {
+        "gene_family": "SPL",
+        "target_species": ["Ft"],
+        "reference_species": ["At"],
+        "stages": authorized,
+        "primary_stage": "combo",
+        "workflow_action_id": "build_subfamily_evidence",
+        "params": {
+            "workflowActionId": "build_subfamily_evidence",
+            "authorizedStages": authorized,
+            "targetSpecies": ["Ft"],
+            "referenceSpecies": ["At"],
+        },
+    }
+    monkeypatch.setattr(
+        api,
+        "_reconcile_confirmed_gene_family_species",
+        lambda *args, **kwargs: {"applicable": False, "conflicts": []},
+    )
+    monkeypatch.setattr(
+        api, "_resolve_authoritative_gene_family", lambda *args, **kwargs: "SPL"
+    )
+    monkeypatch.setattr(
+        api,
+        "_resolve_subfamily_request_species_roles",
+        lambda *args, **kwargs: {
+            "classification_species": ["Ft"],
+            "reference_species": ["At"],
+        },
+    )
+    monkeypatch.setattr(
+        api,
+        "_latest_gene_family_scope_defaults",
+        lambda *args, **kwargs: {
+            "target_species": ["Ft"],
+            "reference_species": ["At"],
+            "comparative_species": [],
+        },
+    )
+    monkeypatch.setattr(
+        api,
+        "_load_identify_species_selection",
+        lambda *args, **kwargs: {
+            "groups": {
+                "target": [{"species": "Ft"}],
+                "reference": [{"species": "At"}],
+                "comparative": [],
+            }
+        },
+    )
+    monkeypatch.setattr(
+        api,
+        "_subfamily_state_for_targets",
+        lambda *args, **kwargs: {"overall": "not_classified", "by_species": {}},
+    )
+    monkeypatch.setattr(
+        api, "_identity_completed_stages_for_plan", lambda *args, **kwargs: []
+    )
+    monkeypatch.setattr(
+        api,
+        "_resolve_contextual_gene_family_request",
+        lambda *args, **kwargs: {
+            "family": "SPL",
+            "prefix": "Ft.SPL",
+            "stages": authorized,
+            "targetSpecies": ["Ft"],
+            "referenceSpecies": ["At"],
+            "comparativeSpecies": [],
+            "scopeMode": "target_plus_reference",
+            "contextGate": {"valid": False, "issues": []},
+            "context_route": "gene_family_default_scope",
+            "requiresConfirmedSubfamily": False,
+            "skipSubfamilyLabels": False,
+            "subfamilyState": {
+                "overall": "not_classified",
+                "by_species": {"Ft": {"status": "not_classified"}},
+            },
+        },
+    )
+
+    response = asyncio.run(
+        api.agent_plan(
+            api.AgentPlanRequest(
+                message="执行接下来的步骤",
+                session_id="authorized-action-test",
+                job_id="job-test",
+                prefix="Ft.SPL",
+                intent=intent,
+            )
+        )
+    )
+
+    expected = ["phylogeny", "motif", "domain", "structure", "combo"]
+    assert response["plan"]["all_stages"] == expected
+    assert response["plan"]["stages"] == expected
+    assert response["intent"]["workflow_action_id"] == "build_subfamily_evidence"
+    assert response["intent"]["params"]["authorizedStages"] == authorized
+    assert "reference_classification" not in response["plan"]["all_stages"]
+    assert "subfamily_confirm" not in response["plan"]["all_stages"]

tests/unit/test_current_recommendation_contract.py

@@ -0,0 +1,450 @@
+import asyncio
+import json
+
+from pydantic import BaseModel
+
+from gfa_core.mas.agent_state_repository import SQLiteAgentStateRepository
+from gfa_core.mas.agents.conversation_tools import ConversationToolRegistry
+from gfa_core.mas.agents.router_runtime import RouterRuntime
+from gfa_core.mas.llm import LLMResponse, LLMToolCall
+from gfa_core.mas.persistent_conversation import PersistentConversationStore
+from gfa_core.mas.workflow_models import CurrentRecommendation, GoalLedger
+
+
+class _NoArgs(BaseModel):
+    pass
+
+
+class _IntentArgs(BaseModel):
+    query: str
+
+
+class _SequenceLLM:
+    def __init__(self, responses):
+        self.responses = list(responses)
+
+    async def acall(self, *_args, **_kwargs):
+        return self.responses.pop(0)
+
+
+def _response(content="", *, tool_calls=None):
+    return LLMResponse(
+        content=content,
+        model="test-model",
+        provider="test-provider",
+        usage={},
+        tool_calls=list(tool_calls or []),
+    )
+
+
+def _recommendation(job_id="job-current-recommendation"):
+    return CurrentRecommendation(
+        recommendation_id="rec-build-subfamily-evidence",
+        job_id=job_id,
+        workflow_id="gene_family",
+        transition_id="build_subfamily_evidence",
+        from_state_id="family_identified",
+        to_state_id="subfamily_evidence_ready",
+        target_stage="combo",
+        authorized_stages=("domain", "motif", "phylogeny", "structure", "combo"),
+        evidence_digest="a" * 64,
+        family="SPL",
+        scope_id="SPL.target-Ft.target_only",
+    )
+
+
+def test_identified_project_context_projects_one_typed_recommendation():
+    recommendation = RouterRuntime._current_workflow_recommendation(
+        "gene_family",
+        "job-project-recommendation",
+        {
+            "family": "SPL",
+            "scopes": [{"scope_id": "SPL.target-Ft.target_only"}],
+            "manifests": [{"stage": "identify", "status": "success"}],
+            "stage_status": {
+                "identify": {"stage": "identify", "status": "success"}
+            },
+        },
+    )
+
+    assert isinstance(recommendation, CurrentRecommendation)
+    assert recommendation.transition_id == "build_subfamily_evidence"
+    assert recommendation.target_stage == "combo"
+    assert recommendation.authorized_stages == (
+        "domain",
+        "motif",
+        "phylogeny",
+        "structure",
+        "combo",
+    )
+
+
+def test_next_analysis_question_stays_read_only_and_returns_typed_recommendation():
+    job_id = "job-next-analysis-question"
+    context = {
+        "job_id": job_id,
+        "family": "SPL",
+        "scopes": [{"scope_id": "SPL.target-Ft.target_only"}],
+        "manifests": [{"stage": "identify", "status": "success"}],
+        "stage_status": {
+            "identify": {"stage": "identify", "status": "success"}
+        },
+    }
+    registry = ConversationToolRegistry()
+    registry.register(
+        name="inspect_project_context",
+        description="Inspect current project context.",
+        args_model=_NoArgs,
+        handler=lambda _args: context,
+        permission="read_only",
+    )
+    llm = _SequenceLLM(
+        [
+            _response(
+                '{"goal":"project_status","needs":["project_context"],'
+                '"subject":"当前工作流的下一业务动作",'
+                '"referent":null,"clarification_question":null,'
+                '"interaction_kind":"status","workflow_id":"gene_family",'
+                '"analysis_request_kind":null,'
+                '"uses_current_recommendation":false,'
+                '"confirms_pending_proposal":false}'
+            ),
+            _response(
+                tool_calls=[
+                    LLMToolCall(
+                        id="inspect-next-analysis-1",
+                        name="inspect_project_context",
+                        arguments="{}",
+                    )
+                ]
+            ),
+            _response(
+                '{"status":"ok","action":"show_result",'
+                '"message":"下一步是构建亚家族证据。"}'
+            ),
+        ]
+    )
+
+    outcome = asyncio.run(
+        RouterRuntime(llm, registry, max_iterations=5).run(
+            query="接下来的分析有哪些？",
+            history=[],
+            session_id="session-next-analysis-question",
+            job_id=job_id,
+            goal_ledger=GoalLedger(
+                conversation_id="conversation-next-analysis-question",
+                revision=0,
+                user_terms={"active_job_id": job_id, "gene_family": "SPL"},
+            ),
+        )
+    )
+
+    assert outcome.action == "show_result"
+    assert outcome.proposal is None
+    assert outcome.recommendation is not None
+    assert outcome.recommendation.transition_id == "build_subfamily_evidence"
+
+
+def test_show_result_persists_and_cancel_clears_current_recommendation(tmp_path):
+    repository = SQLiteAgentStateRepository(tmp_path / "agent-state.sqlite3")
+    repository.initialize()
+    store = PersistentConversationStore(repository)
+    turn = store.begin_turn(
+        session_id="session-current-recommendation",
+        job_id="job-current-recommendation",
+        query="接下来的分析有哪些？",
+        client_history=[],
+    )
+    recommendation = _recommendation().model_dump(mode="json")
+
+    ledger = store.finish_turn(
+        turn,
+        assistant_message="下一步是构建亚家族证据。",
+        interaction_kind="status",
+        action="show_result",
+        recommendation=recommendation,
+    )
+    assert (
+        ledger.model_dump(mode="json")["user_terms"]["current_recommendation"]
+        == recommendation
+    )
+
+    cancel_turn = store.begin_turn(
+        session_id="session-current-recommendation",
+        job_id="job-current-recommendation",
+        query="取消",
+        client_history=[],
+    )
+    ledger = store.finish_turn(
+        cancel_turn,
+        assistant_message="已取消。",
+        interaction_kind="cancel",
+        action="reply",
+    )
+    assert "current_recommendation" not in ledger.user_terms
+
+
+def test_pending_proposal_preserves_complete_workflow_action_binding(tmp_path):
+    repository = SQLiteAgentStateRepository(tmp_path / "agent-state.sqlite3")
+    repository.initialize()
+    store = PersistentConversationStore(repository)
+    turn = store.begin_turn(
+        session_id="session-action-binding",
+        job_id="job-action-binding",
+        query="执行当前推荐",
+        client_history=[],
+    )
+    proposal = {
+        "proposal_id": "proposal-action-binding",
+        "request": "执行 build_subfamily_evidence",
+        "subject": "执行当前推荐",
+        "workflow_id": "gene_family",
+        "workflow_action_id": "build_subfamily_evidence",
+        "authorized_stages": "domain,motif,phylogeny,structure,combo",
+        "target_stage": "combo",
+    }
+
+    ledger = store.finish_turn(
+        turn,
+        assistant_message="请确认。",
+        interaction_kind="supplement",
+        action="clarify",
+        proposal=proposal,
+    )
+
+    assert ledger.user_terms["pending_proposal"] == {
+        **proposal,
+        "job_id": "job-action-binding",
+    }
+
+
+def test_execute_current_recommendation_revalidates_and_limits_intent():
+    job_id = "job-execute-current-recommendation"
+    context = {
+        "job_id": job_id,
+        "family": "SPL",
+        "scopes": [{"scope_id": "SPL.target-Ft.target_only"}],
+        "manifests": [{"stage": "identify", "status": "success"}],
+        "stage_status": {
+            "identify": {"stage": "identify", "status": "success"}
+        },
+    }
+    recommendation = RouterRuntime._current_workflow_recommendation(
+        "gene_family", job_id, context
+    )
+    registry = ConversationToolRegistry()
+    registry.register(
+        name="inspect_project_context",
+        description="Inspect current project context.",
+        args_model=_NoArgs,
+        handler=lambda _args: context,
+        permission="read_only",
+    )
+    registry.register(
+        name="recognize_analysis_intent",
+        description="Recognize an analysis intent.",
+        args_model=_IntentArgs,
+        handler=lambda _args: {
+            "gene_family": "SPL",
+            "stages": ["identify"],
+            "primary_stage": "identify",
+            "params": {"targetSpecies": ["Ft"], "referenceSpecies": ["At"]},
+        },
+        permission="pure",
+    )
+    ledger = GoalLedger(
+        conversation_id="conversation-execute-current-recommendation",
+        revision=0,
+        user_terms={
+            "active_job_id": job_id,
+            "current_recommendation": recommendation.model_dump(mode="json"),
+        },
+    )
+    llm = _SequenceLLM(
+        [
+            _response(
+                '{"goal":"new_analysis","needs":["analysis_intent"],'
+                '"subject":"执行当前推荐动作","referent":"current_recommendation",'
+                '"clarification_question":null,"interaction_kind":"supplement",'
+                '"analysis_request_kind":"execution",'
+                '"uses_current_recommendation":true,'
+                '"confirms_pending_proposal":false}'
+            ),
+            _response(
+                tool_calls=[
+                    LLMToolCall(
+                        id="recognize-current-recommendation-1",
+                        name="recognize_analysis_intent",
+                        arguments=json.dumps(
+                            {"query": "执行接下来的分析"}, ensure_ascii=False
+                        ),
+                    )
+                ]
+            ),
+            _response(
+                '{"status":"ok","action":"propose_analysis",'
+                '"message":"已交给现有分析系统。"}'
+            ),
+        ]
+    )
+
+    outcome = asyncio.run(
+        RouterRuntime(llm, registry, max_iterations=5).run(
+            query="执行接下来的分析",
+            history=[],
+            session_id="session-execute-current-recommendation",
+            job_id=job_id,
+            goal_ledger=ledger,
+        )
+    )
+
+    assert outcome.action == "propose_analysis"
+    assert outcome.intent["stages"] == [
+        "domain",
+        "motif",
+        "phylogeny",
+        "structure",
+        "combo",
+    ]
+    assert outcome.intent["primary_stage"] == "combo"
+    assert outcome.intent["workflow_action_id"] == "build_subfamily_evidence"
+
+
+def test_semantic_business_referent_hands_off_when_workflow_is_terminal():
+    captured_queries = []
+    registry = ConversationToolRegistry()
+    registry.register(
+        name="recognize_analysis_intent",
+        description="Recognize an analysis intent.",
+        args_model=_IntentArgs,
+        handler=lambda args: captured_queries.append(args.query)
+        or {
+            "gene_family": "SPL",
+            "stages": ["promoter", "synteny", "ppi"],
+            "primary_stage": "promoter",
+        },
+        permission="pure",
+    )
+    llm = _SequenceLLM(
+        [
+            _response(
+                '{"goal":"new_analysis","needs":["analysis_intent"],'
+                '"subject":"执行上一轮说明的后续业务",'
+                '"referent":"常规基因家族分析",'
+                '"clarification_question":null,"interaction_kind":"supplement",'
+                '"workflow_id":"gene_family","analysis_request_kind":"execution",'
+                '"uses_current_recommendation":true,'
+                '"confirms_pending_proposal":false}'
+            ),
+            _response(
+                tool_calls=[
+                    LLMToolCall(
+                        id="recognize-semantic-business-referent-1",
+                        name="recognize_analysis_intent",
+                        arguments=json.dumps(
+                            {"query": "执行常规基因家族分析"}, ensure_ascii=False
+                        ),
+                    )
+                ]
+            ),
+            _response(
+                '{"status":"ok","action":"propose_analysis",'
+                '"message":"分析请求已交给现有意图系统。"}'
+            ),
+        ]
+    )
+
+    outcome = asyncio.run(
+        RouterRuntime(llm, registry, max_iterations=5).run(
+            query="执行接下来的分析",
+            history=[
+                {
+                    "role": "assistant",
+                    "content": (
+                        "亚家族分类与确认已经完成，"
+                        "接下来进入常规基因家族分析。"
+                    ),
+                }
+            ],
+            session_id="session-semantic-business-referent",
+            job_id="job-confirmed-family",
+            goal_ledger=GoalLedger(
+                conversation_id="conversation-semantic-business-referent",
+                revision=0,
+                user_terms={"active_job_id": "job-confirmed-family"},
+            ),
+        )
+    )
+
+    assert outcome.action == "propose_analysis"
+    assert captured_queries == ["执行常规基因家族分析"]
+    assert outcome.intent["stages"] == ["promoter", "synteny", "ppi"]
+
+
+def test_unresolved_current_recommendation_without_binding_still_clarifies():
+    llm = _SequenceLLM(
+        [
+            _response(
+                '{"goal":"new_analysis","needs":["analysis_intent"],'
+                '"subject":"执行当前推荐动作",'
+                '"referent":"current_recommendation",'
+                '"clarification_question":null,"interaction_kind":"supplement",'
+                '"workflow_id":"gene_family","analysis_request_kind":"execution",'
+                '"uses_current_recommendation":true,'
+                '"confirms_pending_proposal":false}'
+            )
+        ]
+    )
+
+    outcome = asyncio.run(
+        RouterRuntime(llm, ConversationToolRegistry(), max_iterations=5).run(
+            query="执行接下来的分析",
+            history=[],
+            session_id="session-unresolved-recommendation",
+            job_id="job-unresolved-recommendation",
+            goal_ledger=GoalLedger(
+                conversation_id="conversation-unresolved-recommendation",
+                revision=0,
+                user_terms={"active_job_id": "job-unresolved-recommendation"},
+            ),
+        )
+    )
+
+    assert outcome.action == "clarify"
+    assert "明确" in outcome.message
+
+
+def test_execute_current_recommendation_rejects_cross_job_binding():
+    recommendation = _recommendation(job_id="job-original")
+    ledger = GoalLedger(
+        conversation_id="conversation-cross-job-recommendation",
+        revision=0,
+        user_terms={
+            "current_recommendation": recommendation.model_dump(mode="json")
+        },
+    )
+    llm = _SequenceLLM(
+        [
+            _response(
+                '{"goal":"new_analysis","needs":["analysis_intent"],'
+                '"subject":"执行当前推荐动作","referent":"current_recommendation",'
+                '"clarification_question":null,"interaction_kind":"supplement",'
+                '"analysis_request_kind":"execution",'
+                '"uses_current_recommendation":true,'
+                '"confirms_pending_proposal":false}'
+            )
+        ]
+    )
+
+    outcome = asyncio.run(
+        RouterRuntime(llm, ConversationToolRegistry(), max_iterations=5).run(
+            query="执行接下来的分析",
+            history=[],
+            session_id="session-cross-job-recommendation",
+            job_id="job-other",
+            goal_ledger=ledger,
+        )
+    )
+
+    assert outcome.action == "clarify"
+    assert "当前 job" in outcome.message

tests/unit/test_deepseek_json_protocol.py

@@ -0,0 +1,88 @@
+import asyncio
+from types import SimpleNamespace
+
+from gfa_core.mas.agents.conversation_tools import ConversationToolRegistry
+from gfa_core.mas.agents.router_runtime import RouterRuntime
+from gfa_core.mas.llm import LLMMessage, LLMResponse
+from gfa_core.mas.llm.providers.deepseek import DeepSeekProvider
+
+
+def test_deepseek_provider_forwards_requested_json_mode():
+    captured = {}
+
+    class _Completions:
+        def create(self, **kwargs):
+            captured.update(kwargs)
+            return SimpleNamespace(
+                choices=[
+                    SimpleNamespace(
+                        message=SimpleNamespace(content='{"ok":true}', tool_calls=[]),
+                        finish_reason="stop",
+                    )
+                ],
+                usage=SimpleNamespace(
+                    prompt_tokens=1,
+                    completion_tokens=1,
+                    total_tokens=2,
+                ),
+            )
+
+    provider = DeepSeekProvider(api_key="test-key")
+    provider.client = SimpleNamespace(
+        chat=SimpleNamespace(completions=_Completions())
+    )
+
+    provider.call(
+        [LLMMessage(role="user", content="Return JSON")],
+        response_format={"type": "json_object"},
+    )
+
+    assert captured["response_format"] == {"type": "json_object"}
+
+
+def test_router_route_but_not_freeform_reply_requests_native_json_mode():
+    class _RecordingLLM:
+        def __init__(self):
+            self.calls = []
+            self.responses = [
+                LLMResponse(
+                    content=(
+                        '{"goal":"general","needs":[],"subject":"编写脚本",'
+                        '"referent":null,"clarification_question":null,'
+                        '"interaction_kind":"supplement"}'
+                    ),
+                    model="test-model",
+                    provider="test-provider",
+                    usage={},
+                ),
+                LLMResponse(
+                    content=(
+                        '{"status":"ok","action":"reply",'
+                        '"message":"已生成脚本。"}'
+                    ),
+                    model="test-model",
+                    provider="test-provider",
+                    usage={},
+                ),
+            ]
+
+        async def acall(self, _messages, **kwargs):
+            self.calls.append(kwargs)
+            return self.responses.pop(0)
+
+    llm = _RecordingLLM()
+    outcome = asyncio.run(
+        RouterRuntime(llm, ConversationToolRegistry(), max_iterations=5).run(
+            query="帮我写一个 FASTA 脚本",
+            history=[],
+            session_id="deepseek-json-protocol",
+            job_id="",
+            goal_ledger=None,
+        )
+    )
+
+    assert outcome.action == "reply"
+    assert [call.get("response_format") for call in llm.calls] == [
+        {"type": "json_object"},
+        None,
+    ]

tests/unit/test_gene_family_report.py

@@ -0,0 +1,844 @@
+from __future__ import annotations
+
+import json
+from pathlib import Path
+import re
+
+import pytest
+
+import gfa_core.gene_family_report as report_module
+from gfa_core.gene_family_report import (
+    NoSuccessfulGeneFamilyEvidence,
+    generate_gene_family_report,
+)
+
+
+def _write_manifest(
+    workspace: Path,
+    *,
+    prefix: str,
+    stage: str,
+    status: str = "success",
+    scope_id: str = "scope-a",
+    analysis_route: str = "post_subfamily_confirm",
+    member_set_id: str = "members-a",
+    metric_value: int = 1,
+) -> Path:
+    final_dir = workspace / "final"
+    final_dir.mkdir(parents=True, exist_ok=True)
+    token = f"{scope_id}.{analysis_route}.{member_set_id}"
+    table_path = final_dir / f"{prefix}.{token}.{stage}.tsv"
+    table_path.write_text(f"gene\tvalue\n{scope_id}_gene\t{metric_value}\n", encoding="utf-8")
+    manifest_path = final_dir / f"{prefix}.{token}.{stage}.manifest.json"
+    manifest = {
+        "manifest_version": "1.0",
+        "stage": stage,
+        "status": status,
+        "prefix": prefix,
+        "scope_id": scope_id,
+        "analysis_route": analysis_route,
+        "member_set_id": member_set_id,
+        "parameters": {
+            "scope_id": scope_id,
+            "analysis_route": analysis_route,
+            "member_set_id": member_set_id,
+        },
+        "metrics": {"member_count": metric_value},
+        "artifacts": {
+            "result_table": {
+                "path": f"final/{table_path.name}",
+                "full_path": str(table_path),
+                "type": "table",
+            }
+        },
+    }
+    manifest_path.write_text(json.dumps(manifest, ensure_ascii=False), encoding="utf-8")
+    return manifest_path
+
+
+def _generate(
+    workspace: Path,
+    *,
+    expected_stages: list[str],
+    stage_statuses: dict[str, str],
+    use_model: bool = False,
+    action: str = "generate",
+):
+    return generate_gene_family_report(
+        workspace,
+        "Ft.SPL",
+        expected_stages=expected_stages,
+        stage_statuses=stage_statuses,
+        scope_id="scope-a",
+        analysis_route="post_subfamily_confirm",
+        member_set_id="members-a",
+        use_model=use_model,
+        action=action,
+        job_id="job-report",
+    )
+
+
+def _stage_section(markdown: str, stage: str) -> str:
+    heading = re.search(rf"(?m)^## (?!#).*?\(`{re.escape(stage)}`\)", markdown)
+    if heading is None:
+        raise AssertionError(f"missing report chapter for {stage}")
+    next_heading = re.search(r"(?m)^## (?!#)", markdown[heading.end():])
+    end = heading.end() + next_heading.start() if next_heading else len(markdown)
+    return markdown[heading.start():end]
+
+
+def _subsection(markdown: str, number: str) -> str:
+    heading = re.search(rf"(?m)^### {re.escape(number)}(?:\s|$)", markdown)
+    if heading is None:
+        raise AssertionError(f"missing report subsection {number}")
+    next_heading = re.search(r"(?m)^#{2,3} ", markdown[heading.end():])
+    end = heading.end() + next_heading.start() if next_heading else len(markdown)
+    return markdown[heading.start():end]
+
+
+def _valid_api_arrangement(payload: dict, *, reverse: bool = False) -> dict:
+    sections = []
+    for section in payload["sections"]:
+        sentence_ids = [sentence["sentence_id"] for sentence in section["sentences"]]
+        if reverse:
+            sentence_ids.reverse()
+        sections.append(
+            {
+                "section_id": section["section_id"],
+                "paragraphs": [
+                    {
+                        "sentence_ids": sentence_ids,
+                        "connector_ids": ["also"] * max(0, len(sentence_ids) - 1),
+                    }
+                ],
+            }
+        )
+    return {"sections": sections}
+
+
+def test_report_uses_only_planned_stages_and_current_identity(tmp_path: Path) -> None:
+    workspace = tmp_path / "job"
+    _write_manifest(workspace, prefix="Ft.SPL", stage="identify", metric_value=3)
+    other = _write_manifest(
+        workspace,
+        prefix="Ft.SPL",
+        stage="identify",
+        scope_id="scope-b",
+        analysis_route="target_only",
+        member_set_id="members-b",
+        metric_value=99,
+    )
+
+    result = _generate(
+        workspace,
+        expected_stages=["identify"],
+        stage_statuses={"identify": "success"},
+    )
+
+    report_text = result.report_path.read_text(encoding="utf-8")
+    assert "最终成员数：3" in report_text
+    assert other.name not in report_text
+    generated = {path.name for path in (workspace / "final").glob("Ft.SPL.gene_family_report.*")}
+    assert generated == {
+        "Ft.SPL.gene_family_report.md",
+        "Ft.SPL.gene_family_report.manifest.json",
+    }
+
+
+def test_report_renders_template_subsections_for_only_planned_stages(tmp_path: Path) -> None:
+    workspace = tmp_path / "job"
+    for stage in ("identify", "motif", "pep"):
+        _write_manifest(workspace, prefix="Ft.SPL", stage=stage)
+
+    result = _generate(
+        workspace,
+        expected_stages=["identify", "motif", "pep"],
+        stage_statuses={"identify": "success", "motif": "success", "pep": "success"},
+    )
+
+    markdown = result.report_path.read_text(encoding="utf-8")
+    expected_subsections = {
+        "identify": "### 4.3 最终成员数量与命名",
+        "motif": "### 7.2 Motif 数量与出现频率",
+        "pep": "### 16.1 蛋白理化性质分布图",
+    }
+    for stage, subsection in expected_subsections.items():
+        assert subsection in _stage_section(markdown, stage)
+    assert "(`synteny`)" not in markdown
+    assert "【填写】" not in markdown
+
+
+def test_report_renders_nested_motif_metrics_without_placeholder_values(tmp_path: Path) -> None:
+    workspace = tmp_path / "job"
+    manifest_path = _write_manifest(workspace, prefix="Ft.SPL", stage="motif")
+    manifest = json.loads(manifest_path.read_text(encoding="utf-8"))
+    manifest["metrics"] = {
+        "motif_count": 10,
+        "total_sites": 105,
+        "motifs": [
+            {"width": width}
+            for width in (78, 76, 80, 113, 57, 29, 52, 138, 27, 29)
+        ],
+    }
+    manifest_path.write_text(json.dumps(manifest, ensure_ascii=False), encoding="utf-8")
+
+    result = _generate(
+        workspace,
+        expected_stages=["motif"],
+        stage_statuses={"motif": "success"},
+    )
+
+    motif_section = _stage_section(result.report_path.read_text(encoding="utf-8"), "motif")
+    assert re.search(r"识别\s*10\s*个\s*motif", motif_section)
+    assert re.search(r"总位点数为\s*105", motif_section)
+    assert re.search(r"宽度范围为\s*27–138", motif_section)
+    assert "未解析" not in motif_section
+    assert "N/A" not in motif_section
+    assert "None" not in motif_section
+
+
+def test_identify_uses_real_manifest_member_count_aliases(tmp_path: Path) -> None:
+    workspace = tmp_path / "job"
+    manifest_path = _write_manifest(workspace, prefix="Ft.SPL", stage="identify")
+    manifest = json.loads(manifest_path.read_text(encoding="utf-8"))
+    manifest["metrics"] = {"identified_members_count": 24, "final_members": 24}
+    manifest_path.write_text(json.dumps(manifest, ensure_ascii=False), encoding="utf-8")
+
+    result = _generate(
+        workspace,
+        expected_stages=["identify"],
+        stage_statuses={"identify": "success"},
+    )
+
+    section = _stage_section(result.report_path.read_text(encoding="utf-8"), "identify")
+    assert "最终成员数：24" in section
+    assert re.search(r"形成\s*24\s*个\s*SPL家族", section)
+    assert "源结果未提供此指标" not in section
+
+
+def test_phylogeny_renders_real_manifest_schema_with_explicit_labels(tmp_path: Path) -> None:
+    workspace = tmp_path / "job"
+    manifest_path = _write_manifest(workspace, prefix="Ft.SPL", stage="phylogeny")
+    manifest = json.loads(manifest_path.read_text(encoding="utf-8"))
+    manifest["metrics"] = {"sequences_count": 40, "msa_length": 1449}
+    manifest_path.write_text(json.dumps(manifest, ensure_ascii=False), encoding="utf-8")
+
+    result = _generate(
+        workspace,
+        expected_stages=["phylogeny"],
+        stage_statuses={"phylogeny": "success"},
+    )
+
+    section = _stage_section(result.report_path.read_text(encoding="utf-8"), "phylogeny")
+    assert "序列数：40" in section
+    assert "多序列比对长度：1449" in section
+    assert re.search(r"使用\s*40\s*条序列.*比对长度为\s*1449", section)
+
+
+def test_interspecies_synteny_reads_only_named_real_schema_artifacts(tmp_path: Path) -> None:
+    workspace = tmp_path / "job"
+    manifest_path = _write_manifest(workspace, prefix="Ft.SPL", stage="interspecies_synteny")
+    final_dir = workspace / "final"
+    gene_to_species = final_dir / "Ft.SPL.gene_to_species.tsv"
+    gene_to_species.write_text(
+        "gene_id\tspecies\tgroup\n"
+        "FtGene1.01.T01\tFt\ttarget\n"
+        "FtGene2.01.T01\tFt\ttarget\n"
+        "AT1G00010\tAT\tref\n"
+        "AT1G00020\tAT\tref\n",
+        encoding="utf-8",
+    )
+    collinearity = final_dir / "Ft.At.collinearity"
+    lines = []
+    for index in range(521):
+        lines.append(f"## Alignment {index}: score=1 e_value=1e-10 N=1 Ft&At plus")
+        if index == 0:
+            lines.append("  0-  0:\tAT1G00010.1\tFtGene1.01.T01\t1e-20")
+        elif index == 1:
+            lines.append("  1-  0:\tAT1G00020.1\tFtGene2.01.T01\t1e-20")
+    collinearity.write_text("\n".join(lines) + "\n", encoding="utf-8")
+    manifest = json.loads(manifest_path.read_text(encoding="utf-8"))
+    manifest["metrics"] = {"pair_success_count": 1, "pair_total_count": 1}
+    manifest["artifacts"] = {
+        "gene_to_species": {"path": f"final/{gene_to_species.name}", "type": "tsv"},
+        "Ft.At_collinearity": {"path": f"final/{collinearity.name}", "type": "data"},
+    }
+    manifest_path.write_text(json.dumps(manifest, ensure_ascii=False), encoding="utf-8")
+
+    result = _generate(
+        workspace,
+        expected_stages=["interspecies_synteny"],
+        stage_statuses={"interspecies_synteny": "success"},
+    )
+
+    section = _stage_section(result.report_path.read_text(encoding="utf-8"), "interspecies_synteny")
+    assert "共线性区块数：521" in section
+    assert "家族成员目标—参考配对数：2" in section
+    assert re.search(r"全基因组\s*block\s*521\s*个.*目标-参考家族.*?2\s*对", section)
+
+
+def test_pep_renders_real_manifest_task_summary_with_explicit_labels(tmp_path: Path) -> None:
+    workspace = tmp_path / "job"
+    manifest_path = _write_manifest(workspace, prefix="Ft.SPL", stage="pep")
+    manifest = json.loads(manifest_path.read_text(encoding="utf-8"))
+    manifest["metrics"] = {"protein_count": 40, "tasks_success": 3, "tasks_total": 3}
+    manifest_path.write_text(json.dumps(manifest, ensure_ascii=False), encoding="utf-8")
+
+    result = _generate(
+        workspace,
+        expected_stages=["pep"],
+        stage_statuses={"pep": "success"},
+    )
+
+    section = _stage_section(result.report_path.read_text(encoding="utf-8"), "pep")
+    assert "蛋白数：40" in section
+    assert "子任务成功数：3/3" in section
+    assert re.search(r"覆盖\s*40\s*个蛋白，子任务成功数为\s*3/3", section)
+
+
+def test_report_has_no_empty_third_level_subsections_or_repeated_generic_interpretation(
+    tmp_path: Path,
+) -> None:
+    workspace = tmp_path / "job"
+    for stage in ("identify", "phylogeny", "motif"):
+        manifest_path = _write_manifest(workspace, prefix="Ft.SPL", stage=stage)
+        manifest = json.loads(manifest_path.read_text(encoding="utf-8"))
+        if stage == "phylogeny":
+            manifest["metrics"] = {"sequences_count": 40, "msa_length": 1449}
+        elif stage == "motif":
+            manifest["metrics"] = {
+                "sequence_count": 40,
+                "motif_count": 10,
+                "total_sites": 105,
+                "motifs": [{"width": 27}, {"width": 138}],
+            }
+        manifest_path.write_text(json.dumps(manifest, ensure_ascii=False), encoding="utf-8")
+
+    result = _generate(
+        workspace,
+        expected_stages=["identify", "phylogeny", "motif"],
+        stage_statuses={stage: "success" for stage in ("identify", "phylogeny", "motif")},
+    )
+
+    markdown = result.report_path.read_text(encoding="utf-8")
+    for stage in ("identify", "phylogeny", "motif"):
+        section = _stage_section(markdown, stage)
+        chunks = re.split(r"(?m)(?=^### )", section)[1:]
+        assert chunks
+        for chunk in chunks:
+            body = "\n".join(chunk.splitlines()[1:]).strip()
+            assert body, f"empty subsection: {chunk.splitlines()[0]}"
+    assert markdown.count("本节解释仅限上述可追溯结果，不据此断言真实功能或因果关系。") <= 1
+    assert markdown.count("结果范围受当前输入、参数和源文件完整性限制。") <= 1
+
+
+def test_front_summary_contains_project_numbers_in_six_to_ten_sentences(tmp_path: Path) -> None:
+    workspace = tmp_path / "job"
+    identify_path = _write_manifest(workspace, prefix="Ft.SPL", stage="identify")
+    identify = json.loads(identify_path.read_text(encoding="utf-8"))
+    identify["metrics"] = {"identified_members_count": 24, "final_members": 24}
+    identify_path.write_text(json.dumps(identify, ensure_ascii=False), encoding="utf-8")
+    phylogeny_path = _write_manifest(workspace, prefix="Ft.SPL", stage="phylogeny")
+    phylogeny = json.loads(phylogeny_path.read_text(encoding="utf-8"))
+    phylogeny["metrics"] = {"sequences_count": 40, "msa_length": 1449}
+    phylogeny_path.write_text(json.dumps(phylogeny, ensure_ascii=False), encoding="utf-8")
+
+    result = _generate(
+        workspace,
+        expected_stages=["identify", "phylogeny"],
+        stage_statuses={"identify": "success", "phylogeny": "success"},
+    )
+
+    summary = _subsection(result.report_path.read_text(encoding="utf-8"), "1.3")
+    assert re.search(r"目标物种.*24.*家族成员", summary)
+    assert re.search(r"目标.*参考.*40.*1449", summary)
+    assert 6 <= summary.count("。") <= 10
+
+
+def test_stage_specific_key_facts_land_in_their_semantic_subsections(tmp_path: Path) -> None:
+    workspace = tmp_path / "job"
+    final_dir = workspace / "final"
+    stages = [
+        "reference_classification", "domain", "structure", "combo",
+        "promoter", "tfbs", "synteny",
+    ]
+    manifests = {stage: _write_manifest(workspace, prefix="Ft.SPL", stage=stage) for stage in stages}
+
+    assignment = final_dir / "Ft.SPL.reference_classification.assignment.tsv"
+    assignment.write_text(
+        "target_gene\tassigned_subfamily\tconfidence\n"
+        + "\n".join(
+            f"FtGene{i}\t{'I' if i < 2 else 'II'}\t{confidence}"
+            for i, confidence in enumerate(("high", "high", "medium", "low"), start=1)
+        )
+        + "\n",
+        encoding="utf-8",
+    )
+    reference = json.loads(manifests["reference_classification"].read_text(encoding="utf-8"))
+    reference["metrics"] = {"target_gene_count": 4, "subfamily_rows": 6}
+    reference["artifacts"] = {"assignment_table": {"path": f"final/{assignment.name}", "type": "tsv"}}
+    manifests["reference_classification"].write_text(json.dumps(reference), encoding="utf-8")
+
+    domain = json.loads(manifests["domain"].read_text(encoding="utf-8"))
+    domain["metrics"] = {
+        "proteins_total": 40, "proteins_with_domains": 40, "domain_coverage_rate": 1.0,
+        "top_domains": [["SBP", 40], ["Ig_SPL7", 12]],
+    }
+    manifests["domain"].write_text(json.dumps(domain), encoding="utf-8")
+
+    structure = json.loads(manifests["structure"].read_text(encoding="utf-8"))
+    structure["metrics"] = {
+        "genes_total": 40, "genes_with_structure": 24, "structure_coverage_rate": 0.6,
+        "long_gene_outlier_count": 1,
+        "long_gene_outliers": [{"gene_id": "FtLong1", "length_bp": 12454, "cds_count": 10}],
+    }
+    manifests["structure"].write_text(json.dumps(structure), encoding="utf-8")
+
+    combo = json.loads(manifests["combo"].read_text(encoding="utf-8"))
+    combo["metrics"] = None
+    combo["parameters"] = {"panels": ["motif", "domain", "structure"]}
+    combo["artifacts"] = {
+        "combo_plot_png": {"path": "final/Ft.SPL.combo.png", "type": "image"},
+        "combo_plot_pdf": {"path": "final/Ft.SPL.combo.pdf", "type": "image"},
+    }
+    manifests["combo"].write_text(json.dumps(combo), encoding="utf-8")
+
+    element_counts = final_dir / "Ft.SPL.element_counts.txt"
+    element_counts.write_text("1460\tTATA-box\n852\tCAAT-box\n72\tMYB\n", encoding="utf-8")
+    category_summary = final_dir / "Ft.SPL.promoter_category_summary.tsv"
+    category_summary.write_text("big7\telement_instances\nCore promoter\t2600\nLight response\t188\n", encoding="utf-8")
+    enrichment = final_dir / "Ft.SPL.enrichment_analysis.txt"
+    enrichment.write_text(
+        "Subfamily\tElement\tCount\tSubfamilyTotal\tGlobalCount\tEnrichmentRatio\nI\tAACA_motif\t1\t222\t1\t16.66\n",
+        encoding="utf-8",
+    )
+    cooccurrence = final_dir / "Ft.SPL.cooccurrence_analysis.txt"
+    cooccurrence.write_text(
+        "Element1\tElement2\tCo-occurrenceCount\tGenes\nCAAT-box\tMYB\t23\tFtGene1,FtGene2\n",
+        encoding="utf-8",
+    )
+    complexity = final_dir / "Ft.SPL.complexity_analysis.txt"
+    complexity.write_text(
+        "Gene\tTotalElements\tUniqueElements\tComplexityScore\tComplexityClass\nFtComplex1\t278\t33\t361.44\tVery High\n",
+        encoding="utf-8",
+    )
+    promoter = json.loads(manifests["promoter"].read_text(encoding="utf-8"))
+    promoter["metrics"] = {"total_elements": 3699, "unique_elements": 87, "genes_with_elements": 23}
+    promoter["artifacts"] = {
+        "element_counts": {"path": f"final/{element_counts.name}"},
+        "viz_promoter_position_distribution_summary": {"path": f"final/{category_summary.name}"},
+        "enrichment_analysis": {"path": f"final/{enrichment.name}"},
+        "cooccurrence_analysis": {"path": f"final/{cooccurrence.name}"},
+        "complexity_analysis": {"path": f"final/{complexity.name}"},
+    }
+    manifests["promoter"].write_text(json.dumps(promoter), encoding="utf-8")
+
+    motif_summary = final_dir / "Ft.SPL.tfbs.motif_summary.tsv"
+    motif_summary.write_text(
+        "motif_id\tmotif_alt_id\tfamily_sites\tfamily_genes_with_site\tbackground_sites\tbackground_genes_with_site\n"
+        "MOTIF_A\t\t153\t18\t0\t0\n",
+        encoding="utf-8",
+    )
+    tfbs = json.loads(manifests["tfbs"].read_text(encoding="utf-8"))
+    tfbs["metrics"] = {
+        "family_promoter_count": 24, "background_promoter_count": 0,
+        "family_site_count": 3734, "family_motif_count": 230, "motif_database": "PlantTFDB",
+    }
+    tfbs["artifacts"] = {"motif_summary": {"path": f"final/{motif_summary.name}"}}
+    manifests["tfbs"].write_text(json.dumps(tfbs), encoding="utf-8")
+
+    synteny = json.loads(manifests["synteny"].read_text(encoding="utf-8"))
+    synteny["metrics"] = {"gene_count": 31679, "synteny_blocks": 316, "gene_pairs": 3267, "circos_chromosomes_plotted": 8}
+    manifests["synteny"].write_text(json.dumps(synteny), encoding="utf-8")
+
+    result = _generate(
+        workspace,
+        expected_stages=stages,
+        stage_statuses={stage: "success" for stage in stages},
+    )
+    markdown = result.report_path.read_text(encoding="utf-8")
+    assert re.search(r"4\s*个目标成员.*2\s*个小组.*高信心\s*2", _subsection(markdown, "6.2"))
+    assert re.search(r"40/40.*SBP.*40.*Ig_SPL7.*12", _subsection(markdown, "8.1"))
+    assert re.search(r"24/40.*FtLong1.*12454", _subsection(markdown, "9.1"))
+    assert "Motif、Domain、Structure" in _subsection(markdown, "10.1")
+    promoter_text = _subsection(markdown, "11.1")
+    assert re.search(r"23\s*个目标基因.*3699.*TATA-box.*1460", promoter_text)
+    assert "FtComplex1" in promoter_text and "CAAT-box—MYB" in promoter_text
+    assert re.search(r"24\s*条目标启动子.*3734.*MOTIF_A.*153", _subsection(markdown, "12.1"))
+    assert re.search(r"背景启动子为\s*0.*不能.*富集", _subsection(markdown, "12.1"))
+    assert re.search(r"全基因组.*316.*3267.*8\s*条染色体", _subsection(markdown, "13.1"))
+    assert "源结果未提供家族专属复制基因对" in _subsection(markdown, "13.1")
+
+
+def test_all_planned_success_is_final(tmp_path: Path) -> None:
+    workspace = tmp_path / "job"
+    _write_manifest(workspace, prefix="Ft.SPL", stage="identify")
+    _write_manifest(workspace, prefix="Ft.SPL", stage="motif")
+
+    result = _generate(
+        workspace,
+        expected_stages=["identify", "motif"],
+        stage_statuses={"identify": "success", "motif": "success"},
+    )
+
+    manifest = json.loads(result.manifest_path.read_text(encoding="utf-8"))
+    assert result.report_status == "final"
+    assert manifest["metrics"]["report_status"] == "final"
+    assert manifest["metrics"]["included_stages"] == ["identify", "motif"]
+
+
+def test_unsupported_successful_stage_is_visible_but_not_included_or_final(tmp_path: Path) -> None:
+    workspace = tmp_path / "job"
+    _write_manifest(workspace, prefix="Ft.SPL", stage="identify")
+    _write_manifest(workspace, prefix="Ft.SPL", stage="ppi")
+
+    result = _generate(
+        workspace,
+        expected_stages=["identify", "ppi"],
+        stage_statuses={"identify": "success", "ppi": "success"},
+    )
+
+    markdown = result.report_path.read_text(encoding="utf-8")
+    manifest = json.loads(result.manifest_path.read_text(encoding="utf-8"))
+    assert result.report_status == "partial"
+    assert result.included_stages == ("identify",)
+    assert "ppi" in markdown
+    assert "不支持" in markdown
+    assert "(`ppi`)" not in markdown
+    assert manifest["metrics"]["included_stages"] == ["identify"]
+    assert manifest["metrics"]["unsupported_stages"] == ["ppi"]
+
+
+@pytest.mark.parametrize("expected_stages", [None, []])
+def test_generate_rejects_missing_plan_instead_of_inferring_historical_manifests(
+    tmp_path: Path, expected_stages: list[str] | None
+) -> None:
+    workspace = tmp_path / "job"
+    _write_manifest(workspace, prefix="Ft.SPL", stage="identify")
+
+    with pytest.raises(ValueError, match="expected_stages"):
+        generate_gene_family_report(
+            workspace,
+            "Ft.SPL",
+            expected_stages=expected_stages,
+            stage_statuses={"identify": "success"},
+            scope_id="scope-a",
+            analysis_route="post_subfamily_confirm",
+            member_set_id="members-a",
+            use_model=False,
+            action="generate",
+        )
+
+    assert not list((workspace / "final").glob("*.gene_family_report.*"))
+
+
+def test_report_selects_key_facts_without_dumping_manifest_or_stale_paths(tmp_path: Path) -> None:
+    workspace = tmp_path / "job"
+    manifest_path = _write_manifest(workspace, prefix="Ft.SPL", stage="identify", metric_value=3)
+    manifest = json.loads(manifest_path.read_text(encoding="utf-8"))
+    manifest["metrics"].update({"total_proteins": 31679, "blast_enabled": True})
+    manifest["artifacts"]["result_table"]["path"] = "/home/xzq/old/job/final/Ft.SPL.identify.tsv"
+    manifest["artifacts"]["result_table"]["full_path"] = "/home/xzq/old/job/final/Ft.SPL.identify.tsv"
+    manifest_path.write_text(json.dumps(manifest, ensure_ascii=False), encoding="utf-8")
+
+    result = _generate(
+        workspace,
+        expected_stages=["identify"],
+        stage_statuses={"identify": "success"},
+    )
+
+    markdown = result.report_path.read_text(encoding="utf-8")
+    assert "最终成员数：3" in _subsection(markdown, "4.3")
+    assert not re.search(r"(?m)^- `(?:member_count|total_proteins|blast_enabled)\s*`:", markdown)
+    assert "/home/xzq/" not in markdown
+
+
+def test_fallback_uses_nested_karyotype_metrics(tmp_path: Path) -> None:
+    workspace = tmp_path / "job"
+    manifest_path = _write_manifest(workspace, prefix="Ft.SPL", stage="karyotype")
+    manifest = json.loads(manifest_path.read_text(encoding="utf-8"))
+    manifest["metrics"] = {
+        "species": {
+            "Ft": {
+                "gene_count": 24,
+                "chromosome_count": 8,
+                "chromosomes": [f"Ft{i}" for i in range(1, 9)],
+                "tandem_count": 0,
+            }
+        }
+    }
+    manifest_path.write_text(json.dumps(manifest, ensure_ascii=False), encoding="utf-8")
+
+    result = _generate(
+        workspace,
+        expected_stages=["karyotype"],
+        stage_statuses={"karyotype": "success"},
+    )
+
+    markdown = result.report_path.read_text(encoding="utf-8")
+    assert "24个家族成员定位到8条染色体" in markdown
+    assert "核型分析将N/A" not in markdown
+
+
+def test_artifact_path_remaps_remote_job_to_current_mount(tmp_path: Path) -> None:
+    from gfa_core.mas.agents.interpretation_evidence import _artifact_path
+
+    final_dir = tmp_path / "job" / "final"
+    final_dir.mkdir(parents=True)
+    manifest_path = final_dir / "Ft.SPL.stage.manifest.json"
+    local_artifact = final_dir / "Ft.SPL.result.tsv"
+    local_artifact.write_text("gene\tvalue\nFtSPL1\t1\n", encoding="utf-8")
+    manifest = {
+        "_manifest_path": str(manifest_path),
+        "artifacts": {
+            "result": {
+                "full_path": "/home/xzq/old/PlantAI-dev/workspace/job-old/final/Ft.SPL.result.tsv",
+                "path": "/home/xzq/old/PlantAI-dev/workspace/job-old/final/Ft.SPL.result.tsv",
+            }
+        },
+    }
+
+    assert _artifact_path(manifest, "result") == local_artifact
+
+
+def test_failed_or_missing_planned_stage_is_partial(tmp_path: Path) -> None:
+    workspace = tmp_path / "job"
+    _write_manifest(workspace, prefix="Ft.SPL", stage="identify")
+    _write_manifest(workspace, prefix="Ft.SPL", stage="motif", status="failed")
+
+    result = _generate(
+        workspace,
+        expected_stages=["identify", "motif"],
+        stage_statuses={"identify": "success", "motif": "failed"},
+    )
+
+    markdown = result.report_path.read_text(encoding="utf-8")
+    manifest = json.loads(result.manifest_path.read_text(encoding="utf-8"))
+    assert result.report_status == "partial"
+    assert "motif" in markdown.lower()
+    assert "failed" in markdown.lower()
+    assert manifest["metrics"]["failed_or_missing_stages"] == ["motif"]
+
+
+def test_model_failure_creates_labeled_fallback_report(
+    tmp_path: Path, monkeypatch: pytest.MonkeyPatch
+) -> None:
+    workspace = tmp_path / "job"
+    _write_manifest(workspace, prefix="Ft.SPL", stage="identify")
+
+    def fail_model(_evidence: dict, _stages: list[str]):
+        raise TimeoutError("provider timeout")
+
+    monkeypatch.setattr(report_module, "_generate_api_narrative", fail_model)
+    result = _generate(
+        workspace,
+        expected_stages=["identify"],
+        stage_statuses={"identify": "success"},
+        use_model=True,
+    )
+
+    markdown = result.report_path.read_text(encoding="utf-8")
+    manifest = json.loads(result.manifest_path.read_text(encoding="utf-8"))
+    assert result.generation_mode == "fallback_script"
+    assert "生成模式：候补脚本" in markdown
+    assert manifest["metrics"]["generation_mode"] == "fallback_script"
+    assert manifest["model"]["reason"] == "api_unavailable_or_boundary_violation"
+    assert "{'action':" not in markdown
+
+
+def test_valid_model_sections_create_api_report(
+    tmp_path: Path, monkeypatch: pytest.MonkeyPatch
+) -> None:
+    workspace = tmp_path / "job"
+    _write_manifest(workspace, prefix="Ft.SPL", stage="identify")
+
+    seen_payload: dict = {}
+
+    def valid_model(payload: dict, _stages: list[str]):
+        seen_payload.update(payload)
+        return (
+            _valid_api_arrangement(payload, reverse=True),
+            {"status": "generated", "provider": "test", "model": "test-model"},
+        )
+
+    monkeypatch.setattr(report_module, "_generate_api_narrative", valid_model)
+    result = _generate(
+        workspace,
+        expected_stages=["identify"],
+        stage_statuses={"identify": "success"},
+        use_model=True,
+    )
+
+    markdown = result.report_path.read_text(encoding="utf-8")
+    manifest = json.loads(result.manifest_path.read_text(encoding="utf-8"))
+    assert result.generation_mode == "api"
+    assert set(seen_payload) == {"sections", "allowed_connectors"}
+    assert all(set(section) == {"section_id", "sentences"} for section in seen_payload["sections"])
+    assert all(
+        set(sentence) == {"sentence_id", "text"}
+        for section in seen_payload["sections"]
+        for sentence in section["sentences"]
+    )
+    assert "最终成员数：1" in markdown
+    assert manifest["metrics"]["generation_mode"] == "api"
+    assert "【填写】" not in markdown
+
+
+def test_unknown_model_sentence_id_falls_back_for_the_whole_report(
+    tmp_path: Path, monkeypatch: pytest.MonkeyPatch
+) -> None:
+    workspace = tmp_path / "job"
+    _write_manifest(workspace, prefix="Ft.SPL", stage="identify", metric_value=3)
+
+    def invalid_model(payload: dict, _stages: list[str]):
+        response = _valid_api_arrangement(payload)
+        response["sections"][0]["paragraphs"][0]["sentence_ids"][0] = "unknown.sentence"
+        return response, {"status": "generated"}
+
+    monkeypatch.setattr(report_module, "_generate_api_narrative", invalid_model)
+    result = _generate(
+        workspace,
+        expected_stages=["identify"],
+        stage_statuses={"identify": "success"},
+        use_model=True,
+    )
+
+    markdown = result.report_path.read_text(encoding="utf-8")
+    manifest = json.loads(result.manifest_path.read_text(encoding="utf-8"))
+    assert result.generation_mode == "fallback_script"
+    assert "最终成员数：3" in markdown
+    assert manifest["model"]["reason"] == "api_unavailable_or_boundary_violation"
+
+
+def test_missing_and_duplicate_model_sentence_ids_fall_back_for_the_whole_report(
+    tmp_path: Path, monkeypatch: pytest.MonkeyPatch
+) -> None:
+    workspace = tmp_path / "job"
+    _write_manifest(workspace, prefix="Ft.SPL", stage="identify", metric_value=3)
+
+    def invalid_model(payload: dict, _stages: list[str]):
+        response = _valid_api_arrangement(payload)
+        paragraph = next(
+            item["paragraphs"][0]
+            for item in response["sections"]
+            if len(item["paragraphs"][0]["sentence_ids"]) >= 2
+        )
+        paragraph["sentence_ids"][-1] = paragraph["sentence_ids"][0]
+        return response, {"status": "generated"}
+
+    monkeypatch.setattr(report_module, "_generate_api_narrative", invalid_model)
+    result = _generate(
+        workspace,
+        expected_stages=["identify"],
+        stage_statuses={"identify": "success"},
+        use_model=True,
+    )
+
+    markdown = result.report_path.read_text(encoding="utf-8")
+    manifest = json.loads(result.manifest_path.read_text(encoding="utf-8"))
+    assert result.generation_mode == "fallback_script"
+    assert "最终成员数：3" in markdown
+    assert manifest["model"]["reason"] == "api_unavailable_or_boundary_violation"
+
+
+@pytest.mark.parametrize(
+    "violation",
+    ["unknown_connector", "extra_field", "free_text_field", "changed_section"],
+)
+def test_model_schema_or_connector_boundary_violation_falls_back_for_the_whole_report(
+    tmp_path: Path, monkeypatch: pytest.MonkeyPatch, violation: str
+) -> None:
+    workspace = tmp_path / "job"
+    _write_manifest(workspace, prefix="Ft.SPL", stage="identify", metric_value=3)
+
+    def invalid_model(payload: dict, _stages: list[str]):
+        response = _valid_api_arrangement(payload)
+        if violation == "unknown_connector":
+            paragraph = next(
+                item["paragraphs"][0]
+                for item in response["sections"]
+                if item["paragraphs"][0]["connector_ids"]
+            )
+            paragraph["connector_ids"][0] = "invented"
+        elif violation == "extra_field":
+            response["unexpected"] = True
+        elif violation == "free_text_field":
+            response["sections"][0]["paragraphs"][0]["text"] = "API 自由文本"
+        else:
+            response["sections"][0]["section_id"] = "changed.section"
+        return response, {"status": "generated"}
+
+    monkeypatch.setattr(report_module, "_generate_api_narrative", invalid_model)
+    result = _generate(
+        workspace,
+        expected_stages=["identify"],
+        stage_statuses={"identify": "success"},
+        use_model=True,
+    )
+
+    markdown = result.report_path.read_text(encoding="utf-8")
+    manifest = json.loads(result.manifest_path.read_text(encoding="utf-8"))
+    assert result.generation_mode == "fallback_script"
+    assert "最终成员数：3" in markdown
+    assert "API 自由文本" not in markdown
+    assert manifest["model"]["reason"] == "api_unavailable_or_boundary_violation"
+
+
+def test_open_reuses_existing_report(
+    tmp_path: Path, monkeypatch: pytest.MonkeyPatch
+) -> None:
+    workspace = tmp_path / "job"
+    manifest_path = _write_manifest(workspace, prefix="Ft.SPL", stage="identify")
+    calls = {"count": 0}
+
+    def valid_model(payload: dict, _stages: list[str]):
+        calls["count"] += 1
+        return (
+            _valid_api_arrangement(payload),
+            {"status": "generated", "provider": "test", "model": "test-model"},
+        )
+
+    monkeypatch.setattr(report_module, "_generate_api_narrative", valid_model)
+    first = _generate(
+        workspace,
+        expected_stages=["identify"],
+        stage_statuses={"identify": "success"},
+        use_model=True,
+    )
+    reopened = _generate(
+        workspace,
+        expected_stages=["identify"],
+        stage_statuses={"identify": "success"},
+        use_model=True,
+        action="open",
+    )
+    assert first.reused is False
+    assert reopened.reused is True
+    assert calls["count"] == 1
+
+    manifest = json.loads(manifest_path.read_text(encoding="utf-8"))
+    manifest["metrics"]["member_count"] = 2
+    manifest_path.write_text(json.dumps(manifest, ensure_ascii=False), encoding="utf-8")
+    reopened_after_source_change = _generate(
+        workspace,
+        expected_stages=["identify"],
+        stage_statuses={"identify": "success"},
+        use_model=True,
+        action="open",
+    )
+    assert reopened_after_source_change.reused is True
+    assert calls["count"] == 1
+
+
+def test_no_successful_evidence_creates_no_empty_report(tmp_path: Path) -> None:
+    workspace = tmp_path / "job"
+    _write_manifest(workspace, prefix="Ft.SPL", stage="identify", status="failed")
+
+    with pytest.raises(NoSuccessfulGeneFamilyEvidence, match="暂无可生成"):
+        _generate(
+            workspace,
+            expected_stages=["identify"],
+            stage_statuses={"identify": "failed"},
+        )
+
+    assert not list((workspace / "final").glob("*.gene_family_report.*"))

tests/unit/test_gene_family_report_intent.py

@@ -0,0 +1,112 @@
+from __future__ import annotations
+
+import asyncio
+import json
+
+import pytest
+
+from gfa_core.mas.core.fast_intent import fast_intent_recognition
+from gfa_core.mas.utils.intent_parser import parse_intent_from_query
+
+
+@pytest.mark.parametrize(
+    ("query", "action"),
+    [
+        ("生成结果报告", "generate"),
+        ("生成基因家族结果报告", "generate"),
+        ("打开基因家族结果报告", "open"),
+    ],
+)
+def test_explicit_gene_family_report_intent_is_provider_independent(query: str, action: str) -> None:
+    for parsed in (parse_intent_from_query(query), fast_intent_recognition(query)):
+        assert parsed is not None
+        assert parsed["stages"] == ["gene_family_report"]
+        assert parsed["params"]["geneFamilyReportAction"] == action
+        assert parsed["params"]["reportAction"] == action
+        assert parsed["method"] == "gene_family_report_deterministic"
+
+
+def test_rnaseq_report_intent_is_not_stolen() -> None:
+    for parsed in (
+        parse_intent_from_query("打开 RNA-seq 结果报告"),
+        fast_intent_recognition("打开 RNA-seq 结果报告"),
+    ):
+        assert not parsed or parsed.get("stages") != ["gene_family_report"]
+
+
+@pytest.mark.parametrize(
+    ("query", "action"),
+    [
+        ("生成结果报告", "generate"),
+        ("生成基因家族结果报告", "generate"),
+        ("打开基因家族结果报告", "open"),
+    ],
+)
+def test_gene_family_report_chat_bypasses_router(
+    monkeypatch,
+    tmp_path,
+    query: str,
+    action: str,
+) -> None:
+    from apps.web.vue3 import main as web_main
+    from gfa_core.mas.agents import router as router_module
+
+    final_dir = tmp_path / "workspace" / "job-report" / "final"
+    final_dir.mkdir(parents=True)
+    (final_dir / "Ft.SPL.gene_family_report.manifest.json").write_text(
+        json.dumps(
+            {
+                "stage": "gene_family_report",
+                "status": "success",
+                "prefix": "Ft.SPL",
+                "parameters": {
+                    "expected_stages": ["identify", "motif"],
+                    "stage_statuses": {"identify": "success", "motif": "success"},
+                },
+            }
+        ),
+        encoding="utf-8",
+    )
+
+    async def forbidden_router(*_args, **_kwargs):
+        raise AssertionError("report intent must return before Router/LLM")
+
+    class ForbiddenOrchestrator:
+        async def plan(self, *_args, **_kwargs):
+            raise AssertionError("report plan must not call the LLM orchestrator")
+
+    monkeypatch.setattr(web_main, "PROJECT_ROOT", tmp_path)
+    monkeypatch.setattr(web_main, "WORKSPACE_ROOT", tmp_path / "workspace")
+    monkeypatch.setattr(web_main, "agent_conversation_store", None, raising=False)
+    monkeypatch.setattr(web_main, "llm_manager", object())
+    monkeypatch.setattr(web_main, "orchestrator", ForbiddenOrchestrator())
+    monkeypatch.setattr(web_main, "MAS_AVAILABLE", True)
+    monkeypatch.setattr(router_module.RouterAgent, "execute", forbidden_router)
+
+    result = asyncio.run(
+        web_main._run_agent_chat(
+            web_main.AgentChatRequest(message=query, job_id="job-report")
+        )
+    )
+
+    assert result["action"] == "propose_analysis"
+    assert result["model_used"] == "rule"
+    assert result["intent"]["stages"] == ["gene_family_report"]
+    assert result["intent"]["prefix"] == "Ft.SPL"
+    assert result["intent"]["params"]["reportAction"] == action
+    assert result["intent"]["params"]["expectedStages"] == ["identify", "motif"]
+
+    plan = asyncio.run(
+        web_main.agent_plan(
+            web_main.AgentPlanRequest(
+                message=query,
+                job_id="job-report",
+                prefix="Ft.SPL",
+                intent=result["intent"],
+            )
+        )
+    )
+    assert plan["success"] is True
+    assert plan["mode"] == "gene_family_report_deterministic"
+    assert plan["plan"]["stages"] == ["gene_family_report"]
+    assert plan["plan"]["params"]["expectedStages"] == ["identify", "motif"]

tests/unit/test_gene_family_report_triggers.py

@@ -0,0 +1,181 @@
+from __future__ import annotations
+
+import json
+from pathlib import Path
+
+
+def _request_and_stages():
+    from apps.web.vue3.main import ConfirmedPlanExecutionRequest, ConfirmedPlanStage
+
+    stages = [
+        ConfirmedPlanStage(
+            id="identify",
+            params={
+                "scopeId": "scope-a",
+                "analysisRoute": "post_subfamily_confirm",
+                "memberSetId": "members-a",
+            },
+        ),
+        ConfirmedPlanStage(id="motif", dependencies=["identify"]),
+        ConfirmedPlanStage(id="pep", dependencies=["motif"]),
+    ]
+    request = ConfirmedPlanExecutionRequest(
+        session_id="session-a",
+        job_id="job-a",
+        prefix="Ft.SPL",
+        plan_id="plan-a",
+        stages=stages,
+    )
+    return request, stages
+
+
+def test_confirmed_plan_terminal_success_requests_final_report() -> None:
+    from apps.web.vue3.main import _confirmed_plan_report_request
+
+    request, stages = _request_and_stages()
+    report = _confirmed_plan_report_request(
+        request,
+        stages,
+        {"status": "success", "completed_stages": ["identify", "motif", "pep"]},
+    )
+
+    assert report is not None
+    assert report.stage == "gene_family_report"
+    assert report.expectedStages == ["identify", "motif", "pep"]
+    assert report.stageStatuses == {"identify": "success", "motif": "success", "pep": "success"}
+    assert report.reportAction == "generate"
+
+
+def test_confirmed_plan_failure_requests_partial_report() -> None:
+    from apps.web.vue3.main import _confirmed_plan_report_request
+
+    request, stages = _request_and_stages()
+    report = _confirmed_plan_report_request(
+        request,
+        stages,
+        {
+            "status": "failed",
+            "completed_stages": ["identify"],
+            "failure": {"stage": "motif"},
+        },
+    )
+
+    assert report is not None
+    assert report.stageStatuses == {"identify": "success", "motif": "failed", "pep": "not_run"}
+    assert report.scopeId == "scope-a"
+    assert report.analysisRoute == "post_subfamily_confirm"
+    assert report.memberSetId == "members-a"
+
+
+def test_paused_plan_does_not_auto_finalize_report() -> None:
+    from apps.web.vue3.main import _confirmed_plan_report_request
+
+    request, stages = _request_and_stages()
+    report = _confirmed_plan_report_request(
+        request,
+        stages,
+        {"status": "paused", "completed_stages": ["identify"]},
+    )
+    assert report is None
+
+
+def test_frontend_batch_terminal_path_calls_report_stage_with_planned_statuses() -> None:
+    source = (Path(__file__).resolve().parents[2] / "apps/web/vue3/index.html").read_text(encoding="utf-8")
+
+    assert "submitGeneFamilyReportForBatch" in source
+    assert "executeSingleStageWithParams('gene_family_report'" in source
+    assert "expectedStages: plannedIds" in source
+    assert "stageStatuses" in source
+    assert "if (pausedCount > 0) return" in source
+    assert "continuationOpened" in source
+    assert "name: 'gene_family_report'" in source
+    assert "openGeneratedGeneFamilyReport" in source
+
+
+def test_report_context_prefers_terminal_confirmed_plan(monkeypatch, tmp_path: Path) -> None:
+    from apps.web.vue3 import main as web_main
+
+    final_dir = tmp_path / "job-a" / "final"
+    confirmed_dir = final_dir / ".confirmed_plans"
+    confirmed_dir.mkdir(parents=True)
+    (final_dir / "Ft.SPL.gene_family_report.manifest.json").write_text(
+        json.dumps(
+            {
+                "stage": "gene_family_report",
+                "status": "success",
+                "prefix": "Ft.SPL",
+                "parameters": {
+                    "expected_stages": ["identify"],
+                    "stage_statuses": {"identify": "success"},
+                },
+            }
+        ),
+        encoding="utf-8",
+    )
+    (confirmed_dir / "0123456789abcdefabcd.json").write_text(
+        json.dumps(
+            {
+                "status": "failed",
+                "job_id": "job-a",
+                "prefix": "Ft.SPL",
+                "stages": ["identify", "motif", "pep", "gene_family_report"],
+                "completed_stages": ["identify", "motif"],
+                "failure": {"stage": "pep"},
+                "scopeId": "scope-a",
+                "memberSetId": "members-a",
+            }
+        ),
+        encoding="utf-8",
+    )
+    monkeypatch.setattr(web_main, "WORKSPACE_ROOT", tmp_path)
+
+    context = web_main._gene_family_report_plan_context("job-a")
+
+    assert context == {
+        "prefix": "Ft.SPL",
+        "expectedStages": ["identify", "motif", "pep"],
+        "stageStatuses": {"identify": "success", "motif": "success", "pep": "failed"},
+        "scopeId": "scope-a",
+        "analysisRoute": "",
+        "memberSetId": "members-a",
+        "rnaseqScopeId": "",
+    }
+
+
+def test_report_context_reuses_formal_report_plan_only(monkeypatch, tmp_path: Path) -> None:
+    from apps.web.vue3 import main as web_main
+
+    final_dir = tmp_path / "job-a" / "final"
+    final_dir.mkdir(parents=True)
+    (final_dir / "Ft.SPL.motif.manifest.json").write_text(
+        json.dumps({"stage": "motif", "status": "success", "prefix": "Ft.SPL"}),
+        encoding="utf-8",
+    )
+    monkeypatch.setattr(web_main, "WORKSPACE_ROOT", tmp_path)
+    assert web_main._gene_family_report_plan_context("job-a") is None
+
+    (final_dir / "Ft.SPL.gene_family_report.manifest.json").write_text(
+        json.dumps(
+            {
+                "stage": "gene_family_report",
+                "status": "success",
+                "prefix": "Ft.SPL",
+                "scope_id": "scope-old",
+                "parameters": {
+                    "expected_stages": ["identify", "motif"],
+                    "stage_statuses": {"identify": "success", "motif": "success"},
+                },
+            }
+        ),
+        encoding="utf-8",
+    )
+
+    context = web_main._gene_family_report_plan_context("job-a", "Ft.SPL")
+
+    assert context is not None
+    assert context["expectedStages"] == ["identify", "motif"]
+    assert context["stageStatuses"] == {"identify": "success", "motif": "success"}
+    assert context["scopeId"] == ""
+    assert context["analysisRoute"] == ""
+    assert context["memberSetId"] == ""
+    assert context["rnaseqScopeId"] == ""

tests/unit/test_gene_family_report_wiring.py

@@ -0,0 +1,347 @@
+from __future__ import annotations
+
+import json
+from pathlib import Path
+import subprocess
+import sys
+
+
+def _write_identify_manifest(workspace: Path, prefix: str = "Ft.SPL") -> Path:
+    final_dir = workspace / "final"
+    final_dir.mkdir(parents=True, exist_ok=True)
+    table = final_dir / f"{prefix}.identify.tsv"
+    table.write_text("gene\tvalue\nFtSPL1\t1\n", encoding="utf-8")
+    manifest = final_dir / f"{prefix}.identify.manifest.json"
+    manifest.write_text(
+        json.dumps(
+            {
+                "manifest_version": "1.0",
+                "stage": "identify",
+                "status": "success",
+                "prefix": prefix,
+                "metrics": {"member_count": 1},
+                "artifacts": {
+                    "members": {
+                        "path": f"final/{table.name}",
+                        "full_path": str(table),
+                        "type": "table",
+                    }
+                },
+            },
+            ensure_ascii=False,
+        ),
+        encoding="utf-8",
+    )
+    return manifest
+
+
+def _flag_value(command: list[str], flag: str) -> str:
+    return command[command.index(flag) + 1]
+
+
+def test_report_stage_is_registered_without_analysis_dependencies() -> None:
+    from gfa_core.mas.stage_capabilities import CAPABILITIES
+    from gfa_core.mas.stage_catalog import STAGE_ORDER, is_stage_executable
+    from gfa_core.mas.tools.stage_tools import create_stage_registry
+
+    capability = CAPABILITIES["gene_family_report"]
+    stage_tool = create_stage_registry().get("gene_family_report")
+
+    assert "gene_family_report" in STAGE_ORDER
+    assert is_stage_executable("gene_family_report")
+    assert capability.consumes == []
+    assert stage_tool is not None
+    assert stage_tool.required_inputs == []
+    assert stage_tool.script == "run_gene_family_report.py"
+
+
+def test_report_cli_can_start_as_a_script() -> None:
+    root = Path(__file__).resolve().parents[2]
+    completed = subprocess.run(
+        [sys.executable, "apps/cli/run_gene_family_report.py", "--help"],
+        cwd=root,
+        capture_output=True,
+        text=True,
+        check=False,
+    )
+
+    assert completed.returncode == 0, completed.stderr
+
+
+def test_report_cli_writes_report_and_manifest(tmp_path: Path) -> None:
+    from apps.cli.run_gene_family_report import main
+
+    workspace = tmp_path / "job"
+    _write_identify_manifest(workspace)
+    result = main(
+        [
+            "--workspace",
+            str(workspace),
+            "--prefix",
+            "Ft.SPL",
+            "--job-id",
+            "job-wiring",
+            "--expected-stages",
+            "identify",
+            "--stage-statuses-json",
+            '{"identify":"success"}',
+            "--no-model",
+        ]
+    )
+
+    manifest = json.loads(result.manifest_path.read_text(encoding="utf-8"))
+    assert result.report_path.is_file()
+    assert result.manifest_path.is_file()
+    assert "gene_family_report" in manifest["artifacts"]
+
+
+def test_pipeline_backend_passes_plan_status_and_identity_to_report_cli(
+    tmp_path: Path, monkeypatch
+) -> None:
+    from apps.web import pipeline_backend
+
+    workspace = tmp_path / "job"
+    input_dir = workspace / "input"
+    logs_dir = workspace / "logs"
+    input_dir.mkdir(parents=True)
+    logs_dir.mkdir(parents=True)
+    _write_identify_manifest(workspace)
+    captured: dict[str, list[str]] = {}
+
+    def capture(command, _log_file, cwd=None):
+        captured["command"] = [str(value) for value in command]
+        return 0
+
+    monkeypatch.setattr(pipeline_backend, "can_start_new_task", lambda: (True, ""))
+    monkeypatch.setattr(pipeline_backend, "run_cmd_logged", capture)
+    job = pipeline_backend.JobInfo(
+        job_id="job-backend",
+        input_dir=str(input_dir),
+        workspace_dir=str(workspace),
+        logs_dir=str(logs_dir),
+        created_at="2026-08-08 00:00:00",
+    )
+
+    pipeline_backend.run_stage(
+        "gene_family_report",
+        job,
+        "Ft.SPL",
+        expected_stages=["identify", "motif"],
+        stage_statuses={"identify": "success", "motif": "failed"},
+        report_action="open",
+        report_use_model=False,
+        scope_id="scope-a",
+        analysis_route="post_subfamily_confirm",
+        member_set_id="members-a",
+        rnaseq_scope_id="rna-a",
+    )
+
+    command = captured["command"]
+    assert command[1:3] == ["apps/cli/run_gene_family_report.py", "--workspace"]
+    assert _flag_value(command, "--expected-stages") == "identify,motif"
+    assert json.loads(_flag_value(command, "--stage-statuses-json")) == {
+        "identify": "success",
+        "motif": "failed",
+    }
+    assert _flag_value(command, "--action") == "open"
+    assert _flag_value(command, "--job-id") == "job-backend"
+    assert _flag_value(command, "--scope-id") == "scope-a"
+    assert _flag_value(command, "--analysis-route") == "post_subfamily_confirm"
+    assert _flag_value(command, "--member-set-id") == "members-a"
+    assert _flag_value(command, "--rnaseq-scope-id") == "rna-a"
+    assert "--no-model" in command
+    assert not any("run_identify.py" in value or "run_motif.py" in value for value in command)
+
+
+def test_pipeline_backend_rejects_empty_plan_before_starting_report_cli(
+    tmp_path: Path, monkeypatch
+) -> None:
+    from apps.web import pipeline_backend
+
+    workspace = tmp_path / "job"
+    input_dir = workspace / "input"
+    logs_dir = workspace / "logs"
+    input_dir.mkdir(parents=True)
+    logs_dir.mkdir(parents=True)
+    invoked = False
+
+    def capture(*_args, **_kwargs):
+        nonlocal invoked
+        invoked = True
+        return 0
+
+    monkeypatch.setattr(pipeline_backend, "can_start_new_task", lambda: (True, ""))
+    monkeypatch.setattr(pipeline_backend, "run_cmd_logged", capture)
+    job = pipeline_backend.JobInfo(
+        job_id="job-empty-plan",
+        input_dir=str(input_dir),
+        workspace_dir=str(workspace),
+        logs_dir=str(logs_dir),
+        created_at="2026-08-08 00:00:00",
+    )
+
+    message, _manifest = pipeline_backend.run_stage(
+        "gene_family_report",
+        job,
+        "Ft.SPL",
+        expected_stages=[],
+        report_action="generate",
+        report_use_model=False,
+    )
+
+    assert "expected_stages" in message
+    assert invoked is False
+
+
+def test_stage_tool_executor_forwards_gene_family_report_context(
+    tmp_path: Path, monkeypatch
+) -> None:
+    import asyncio
+    import types
+    from dataclasses import dataclass
+
+    from gfa_core.mas.tools.stage_tools import StageTool, StageToolExecutor
+
+    captured = {}
+
+    @dataclass
+    class FakeJobInfo:
+        job_id: str
+        input_dir: str
+        workspace_dir: str
+        logs_dir: str
+        created_at: str
+
+    def fake_run_stage(**kwargs):
+        captured.update(kwargs)
+        manifest = Path(kwargs["job"].workspace_dir) / "final" / "Ft.SPL.gene_family_report.manifest.json"
+        manifest.parent.mkdir(parents=True, exist_ok=True)
+        manifest.write_text(
+            json.dumps({"stage": "gene_family_report", "status": "success"}),
+            encoding="utf-8",
+        )
+        return "报告已生成", manifest
+
+    fake_backend = types.ModuleType("apps.web.pipeline_backend")
+    fake_backend.JobInfo = FakeJobInfo
+    fake_backend.run_stage = fake_run_stage
+    monkeypatch.setitem(sys.modules, "apps.web.pipeline_backend", fake_backend)
+
+    result = asyncio.run(
+        StageToolExecutor(project_root=tmp_path).execute(
+            StageTool(
+                "gene_family_report",
+                "Gene Family Report",
+                "run_gene_family_report.py",
+                "report",
+                "test",
+                timeout=5,
+            ),
+            "job-report",
+            "Ft.SPL",
+            {
+                "expectedStages": ["identify", "motif"],
+                "stageStatuses": {"identify": "success", "motif": "failed"},
+                "reportAction": "open",
+                "reportUseModel": False,
+                "scopeId": "scope-a",
+                "rnaseqScopeId": "rna-a",
+            },
+        )
+    )
+
+    assert result["status"] == "success"
+    assert captured["expected_stages"] == ["identify", "motif"]
+    assert captured["stage_statuses"] == {"identify": "success", "motif": "failed"}
+    assert captured["report_action"] == "open"
+    assert captured["report_use_model"] is False
+    assert captured["scope_id"] == "scope-a"
+    assert captured["rnaseq_scope_id"] == "rna-a"
+
+
+def test_execute_report_does_not_invent_scope_identity(monkeypatch) -> None:
+    import asyncio
+    from unittest.mock import Mock
+
+    from apps.web.vue3 import main as web_main
+    from gfa_core.mas import resource_governor
+    from tasks import execute_single_stage_task
+
+    submitted = Mock()
+    submitted.return_value.id = "report-task"
+    monkeypatch.setattr(execute_single_stage_task, "apply_async", submitted)
+    monkeypatch.setattr(
+        web_main,
+        "_ensure_celery_queue_health",
+        lambda *_args, **_kwargs: ({"available": True}, None),
+    )
+    monkeypatch.setattr(
+        web_main,
+        "_ensure_gene_family_scope_file",
+        lambda **_kwargs: ("", ""),
+    )
+    monkeypatch.setattr(resource_governor, "record_task", lambda **_kwargs: {})
+
+    request = web_main.ExecuteRequest(
+        stage="gene_family_report",
+        job_id="job-report-no-invented-identity",
+        prefix="Ft.SPL",
+        expectedStages=["identify", "motif"],
+        stageStatuses={"identify": "success", "motif": "success"},
+    )
+    result = asyncio.run(web_main.execute_stage(request))
+
+    assert result["success"] is True
+    params = submitted.call_args.kwargs["kwargs"]["params"]
+    assert params["scopeId"] == ""
+    assert params["analysisRoute"] == ""
+    assert params["memberSetId"] == ""
+    assert params["rnaseqScopeId"] == ""
+
+
+def test_report_artifact_validation_does_not_require_invented_identity(monkeypatch) -> None:
+    from types import SimpleNamespace
+
+    import tasks
+    from gfa_core.mas.tools import stage_identity
+
+    captured = {}
+    identity = SimpleNamespace(
+        scope_id="",
+        analysis_route="custom",
+        member_set_id="Ft.SPL:custom:all_members",
+        wgs_sample_set_id="",
+        wgs_reference_id="",
+        wgs_analysis_id="",
+        genotype_scope_id="",
+        phenotype_scope_id="",
+        covariate_scope_id="",
+        model_scope_id="",
+        source_model_scope_id="",
+        analysis_scope_id="",
+    )
+    monkeypatch.setattr(stage_identity, "resolve_stage_identity", lambda **_kwargs: identity)
+
+    def capture_status(**kwargs):
+        captured.update(kwargs)
+        return SimpleNamespace(completed=True)
+
+    monkeypatch.setattr(stage_identity, "get_stage_status_for_identity", capture_status)
+
+    failure = tasks.validate_single_stage_artifacts(
+        "job-report",
+        "Ft.SPL",
+        "gene_family_report",
+        {
+            "manifest": {
+                "stage": "gene_family_report",
+                "status": "success",
+                "prefix": "Ft.SPL",
+                "analysis_route": "",
+                "member_set_id": "",
+            }
+        },
+    )
+
+    assert failure is None
+    assert captured["strict_identity"] is False

tests/unit/test_intent_v2_fallback_safety.py

@@ -0,0 +1,70 @@
+import asyncio
+from types import SimpleNamespace
+
+from gfa_core.mas.agents.base import AgentMessage, AgentState
+from gfa_core.mas.agents.intent_v2 import IntentAgentV2
+
+
+def test_fallback_does_not_turn_unrecognized_continuation_into_identify():
+    intent = IntentAgentV2(None)._fallback_recognition("执行接下来的分析")
+
+    assert intent["stages"] == []
+    assert intent["needs_clarification"] is True
+    assert intent["clarification_type"] == "analysis_stage"
+
+
+def test_validation_does_not_turn_empty_model_stages_into_identify():
+    intent = IntentAgentV2(None)._validate_intent(
+        {
+            "gene_family": "SPL",
+            "species": "Ft",
+            "stages": [],
+            "params": {},
+        },
+        "执行接下来的步骤",
+    )
+
+    assert intent["stages"] == []
+    assert intent["needs_clarification"] is True
+    assert intent["clarification_type"] == "analysis_stage"
+
+
+def test_fallback_still_recognizes_explicit_identify():
+    intent = IntentAgentV2(None)._fallback_recognition("执行 identify 分析")
+
+    assert intent["stages"] == ["identify"]
+    assert intent.get("needs_clarification") is not True
+
+
+def test_empty_model_response_cannot_authorize_identify_and_keeps_goal_context():
+    class _EmptyLLM:
+        async def acall(self, *_args, **_kwargs):
+            return SimpleNamespace(content="")
+
+    state = AgentState(session_id="empty-intent", job_id="job-empty-intent")
+    message = AgentMessage(
+        sender="router",
+        receiver="intent_v2",
+        message_type="recognize_intent",
+        payload={
+            "query": "执行接下来的分析",
+            "job_id": "job-empty-intent",
+            "workspace": "/tmp/job-empty-intent",
+            "goal_context": {
+                "user_terms": {
+                    "gene_family": "SPL",
+                    "target_species": ["Ft"],
+                    "reference_species": ["At"],
+                    "comparative_species": [],
+                }
+            },
+        },
+    )
+
+    returned = asyncio.run(IntentAgentV2(_EmptyLLM()).execute(state, message))
+    intent = returned.results["results"]["intent"]
+
+    assert intent["stages"] == []
+    assert intent["needs_clarification"] is True
+    assert intent["target_species"] == ["Ft"]
+    assert intent["reference_species"] == ["At"]

tests/unit/test_main_agent_data_preparation_handoff.py

@@ -0,0 +1,85 @@
+import json
+
+from gfa_core.mas.agents.conversation_tools import build_plantai_conversation_tools
+from gfa_core.mas.llm import LLMManager
+from gfa_core.mas.workflow_models import GoalLedger
+
+
+def _registry(tmp_path, *, job_id="job-data-preparation-handoff"):
+    ledger = GoalLedger(
+        conversation_id="conversation-data-preparation-handoff",
+        revision=0,
+        user_terms={
+            "active_job_id": job_id,
+            "gene_family": "SPL",
+            "target_species": ["Ft"],
+            "reference_species": ["At"],
+            "config_confirmed": True,
+        },
+    )
+    return build_plantai_conversation_tools(
+        llm_manager=LLMManager({}),
+        project_root=tmp_path,
+        bound_job_id=job_id,
+        goal_ledger=ledger,
+    )
+
+
+def test_analysis_handoff_is_not_rewritten_to_genome_prepare(tmp_path):
+    resolved = _registry(tmp_path).resolve_analysis_handoff_request(
+        "执行亚家族分类",
+        workflow_id="gene_family",
+        guidance=True,
+    )
+
+    assert resolved.startswith("执行亚家族分类\n")
+    assert "执行数据准备（genome_prepare）" not in resolved
+    assert "目标物种 Ft" in resolved
+    assert "参考物种 At" in resolved
+    assert "正式 config.xlsx 已确认" in resolved
+
+
+def test_genome_prepare_manifest_does_not_change_analysis_handoff(tmp_path):
+    job_id = "job-manifest-independent-handoff"
+    registry = _registry(tmp_path, job_id=job_id)
+    before = registry.resolve_analysis_handoff_request(
+        "执行亚家族分类",
+        workflow_id="gene_family",
+        guidance=True,
+    )
+    final_dir = tmp_path / "workspace" / job_id / "final"
+    final_dir.mkdir(parents=True)
+    (final_dir / "genome_prepare.genome_prepare.manifest.json").write_text(
+        json.dumps({"stage": "genome_prepare", "status": "success"}),
+        encoding="utf-8",
+    )
+
+    after = registry.resolve_analysis_handoff_request(
+        "执行亚家族分类",
+        workflow_id="gene_family",
+        guidance=True,
+    )
+
+    assert after == before
+
+
+def test_explicit_data_preparation_request_is_preserved(tmp_path):
+    resolved = _registry(tmp_path).resolve_analysis_handoff_request(
+        "执行数据准备",
+        workflow_id="data_preparation",
+        guidance=False,
+    )
+
+    assert resolved.startswith("执行数据准备\n")
+    assert "当前业务动作已由数据准备契约唯一确定" not in resolved
+
+
+def test_next_step_question_is_not_rewritten_to_genome_prepare(tmp_path):
+    resolved = _registry(tmp_path).resolve_analysis_handoff_request(
+        "接下来需要做什么？",
+        workflow_id="gene_family",
+        guidance=True,
+    )
+
+    assert resolved.startswith("接下来需要做什么？\n")
+    assert "执行数据准备（genome_prepare）" not in resolved

tests/unit/test_main_agent_ordinary_drafting_continuation.py

@@ -0,0 +1,204 @@
+import asyncio
+
+from pydantic import BaseModel
+
+from gfa_core.mas.agents.conversation_tools import ConversationToolRegistry
+from gfa_core.mas.agents.router_runtime import RouterRuntime
+from gfa_core.mas.llm import LLMResponse, LLMToolCall
+
+
+_HISTORY = [
+    {"role": "user", "content": "如果我想写个脚本来修订我的蛋白id"},
+    {"role": "assistant", "content": "请提供输入格式、修订规则和输出要求。"},
+    {"role": "user", "content": ">id.02020202 | name:123 | xxx"},
+    {"role": "assistant", "content": "请确认修订规则。"},
+    {"role": "user", "content": "仅保留第一个字段（id.02020202）"},
+    {
+        "role": "assistant",
+        "content": "请确认输入文件格式，以及覆盖原文件还是输出到新文件？",
+    },
+]
+
+
+def _response(content: str) -> LLMResponse:
+    return LLMResponse(
+        content=content,
+        model="test-model",
+        provider="test-provider",
+        usage={},
+    )
+
+
+class _SequenceLLM:
+    def __init__(self, responses):
+        self.responses = list(responses)
+        self.calls = 0
+
+    async def acall(self, _messages, **_kwargs):
+        self.calls += 1
+        return self.responses.pop(0)
+
+
+class _QueryArgs(BaseModel):
+    query: str
+
+
+def test_drafting_profile_preserves_short_specification_follow_ups():
+    messages = RouterRuntime._route_messages(
+        query="FASTA格式，输出到新文件",
+        history=_HISTORY,
+        session_id="drafting-continuation",
+        job_id="job-bound-to-gene-family",
+        goal_ledger=None,
+    )
+    prompt = " ".join(messages[0].content.split())
+
+    assert "Multi-turn drafting continuity" in prompt
+    assert "a supplement to the same general goal" in prompt
+    assert "must not be routed to utility" in prompt
+    assert "actually transform a current or supplied file" in prompt
+
+
+def test_explicit_script_request_cannot_be_promoted_to_utility_by_router():
+    llm = _SequenceLLM(
+        [
+            _response(
+                '{"goal":"utility","needs":["utility_intent"],'
+                '"subject":"简化FASTA ID","referent":"test.fa",'
+                '"clarification_question":null,"interaction_kind":"independent"}'
+            ),
+            _response(
+                '{"status":"ok","action":"reply",'
+                '"message":"下面给出一个只生成、不执行的 Python 脚本。"}'
+            ),
+        ]
+    )
+
+    outcome = asyncio.run(
+        RouterRuntime(llm, ConversationToolRegistry(), max_iterations=5).run(
+            query="帮我写脚本简化 test.fa 的id",
+            history=[],
+            session_id="explicit-script-request",
+            job_id="",
+            goal_ledger=None,
+        )
+    )
+
+    assert outcome.action == "reply"
+    assert outcome.tool_trace == []
+    assert llm.calls == 2
+
+
+def test_drafting_follow_up_uses_responder_instead_of_router_clarification():
+    llm = _SequenceLLM(
+        [
+            _response(
+                '{"goal":"general","needs":[],"subject":"补充FASTA脚本规则",'
+                '"referent":"蛋白ID脚本",'
+                '"clarification_question":"请重新提供全部需求。",'
+                '"interaction_kind":"supplement"}'
+            ),
+            _response(
+                '{"status":"ok","action":"reply",'
+                '"message":"已记录保留第一个字段；还需要确认输入格式和输出方式。"}'
+            ),
+        ]
+    )
+
+    outcome = asyncio.run(
+        RouterRuntime(llm, ConversationToolRegistry(), max_iterations=5).run(
+            query="仅保留第一个字段（id.02020202）",
+            history=_HISTORY[:4],
+            session_id="drafting-follow-up",
+            job_id="",
+            goal_ledger=None,
+        )
+    )
+
+    assert outcome.action == "reply"
+    assert outcome.message.startswith("已记录保留第一个字段")
+    assert llm.calls == 2
+
+
+def test_locked_general_reply_accepts_bare_drafting_text_without_retry():
+    llm = _SequenceLLM(
+        [
+            _response(
+                '{"goal":"general","needs":[],"subject":"生成FASTA脚本",'
+                '"referent":null,"clarification_question":null,'
+                '"interaction_kind":"supplement"}'
+            ),
+            _response("```python\nprint('FASTA')\n```"),
+            _response(
+                '{"status":"ok","action":"reply",'
+                '"message":"recovered after retry"}'
+            ),
+        ]
+    )
+
+    outcome = asyncio.run(
+        RouterRuntime(llm, ConversationToolRegistry(), max_iterations=5).run(
+            query="FASTA格式，输出到新文件",
+            history=_HISTORY,
+            session_id="bare-drafting-reply",
+            job_id="",
+            goal_ledger=None,
+        )
+    )
+
+    assert outcome.action == "reply"
+    assert outcome.message == "```python\nprint('FASTA')\n```"
+    assert outcome.tool_trace == []
+    assert llm.calls == 2
+
+
+def test_unrecognized_file_utility_returns_actionable_chinese_clarification():
+    registry = ConversationToolRegistry()
+    registry.register(
+        name="recognize_utility_intent",
+        description="Recognize a supported utility request.",
+        args_model=_QueryArgs,
+        handler=lambda _args: {
+            "success": False,
+            "error_code": "utility_intent_not_found",
+            "message": "No supported utility request was recognized.",
+        },
+        permission="pure",
+    )
+    llm = _SequenceLLM(
+        [
+            _response(
+                '{"goal":"utility","needs":["utility_intent"],'
+                '"subject":"处理文件","referent":null,'
+                '"clarification_question":null,"interaction_kind":"independent"}'
+            ),
+            LLMResponse(
+                content="",
+                model="test-model",
+                provider="test-provider",
+                usage={},
+                tool_calls=[
+                    LLMToolCall(
+                        id="utility-not-found-1",
+                        name="recognize_utility_intent",
+                        arguments='{"query":"帮我处理一下文件"}',
+                    )
+                ],
+            ),
+        ]
+    )
+
+    outcome = asyncio.run(
+        RouterRuntime(llm, registry, max_iterations=5).run(
+            query="帮我处理一下文件",
+            history=[],
+            session_id="unrecognized-utility",
+            job_id="job-utility",
+            goal_ledger=None,
+        )
+    )
+
+    assert outcome.action == "clarify"
+    assert "实际文件路径" in outcome.message
+    assert "处理规则" in outcome.message
+    assert "I couldn't identify" not in outcome.message

tests/unit/test_main_agent_product_business_knowledge.py

@@ -0,0 +1,113 @@
+from gfa_core.mas.agents.router_runtime import RouteDecision, RouterRuntime
+
+
+def _general_prompt(query: str, history=None) -> str:
+    return RouterRuntime._initial_messages(
+        query=query,
+        history=list(history or []),
+        session_id="product-business-session",
+        job_id="job-bound-to-other-workflow",
+        decision=RouteDecision(
+            goal="general",
+            needs=[],
+            subject="解释产品业务",
+            interaction_kind="independent",
+        ),
+        goal_ledger=None,
+    )[0].content
+
+
+def _route_prompt(query: str) -> str:
+    return RouterRuntime._route_messages(
+        query=query,
+        history=[],
+        session_id="product-business-route",
+        job_id="job-bound-to-other-workflow",
+        goal_ledger=None,
+    )[0].content
+
+
+def test_gene_family_product_help_defines_two_level_business_view():
+    prompt = _general_prompt("基因家族有哪些分析步骤？")
+    compact = " ".join(prompt.split())
+
+    assert "亚家族分类与确认（前置）" in prompt
+    assert "常规基因家族分析（主体）" in prompt
+    assert "First answer with these two parts only" in prompt
+    assert "Do not enumerate any individual stage in that first answer" in compact
+    assert "must not use RAG or project tools" in compact
+    assert (
+        "phylogeny,motif,domain,structure,combo,promoter,tfbs,synteny,"
+        "interspecies_synteny,karyotype,pep,ortholog,kaks,ppi"
+    ) in prompt
+
+
+def test_rnaseq_product_help_defines_modes_and_authoritative_inputs():
+    prompt = _general_prompt("系统中 RNA-seq 分析有哪些步骤？")
+    compact = " ".join(prompt.split())
+
+    assert "全基因组 RNA-seq 分析（主体）" in prompt
+    assert "已确认基因家族成员的 RNA 结果投影" in prompt
+    assert "First answer with these two modes only" in compact
+    assert "config.xlsx and samples.yaml are equally authoritative inputs" in compact
+    assert (
+        "rna_fact_projection,rnaseq_qc,rnaseq_align,rnaseq_quant,"
+        "rnaseq_sample_qc,rnaseq_de"
+    ) in prompt
+    assert "must not rerun a separate family-mode DE" in compact
+
+
+def test_rnaseq_product_question_and_execution_have_distinct_route_examples():
+    prompt = _route_prompt("全基因组 RNA-seq 分析的正式输入和具体步骤是什么？")
+
+    assert (
+        '"全基因组 RNA-seq 分析的正式输入和具体步骤是什么？" -> '
+        "general + needs=[]."
+    ) in prompt
+    assert (
+        '"执行全基因组 RNA-seq 分析" -> new_analysis + analysis_intent.'
+    ) in prompt
+
+
+def test_product_follow_up_keeps_all_prior_business_context():
+    history = [
+        {"role": "user", "content": "系统中转录组分析有哪些步骤？"},
+        {
+            "role": "assistant",
+            "content": "分为全基因组 RNA-seq 和已确认家族成员的 RNA 结果投影。",
+        },
+        {"role": "user", "content": "先说全基因组这一部分。"},
+        {
+            "role": "assistant",
+            "content": "需要先确认 config.xlsx 和 samples.yaml。",
+        },
+    ]
+    messages = RouterRuntime._initial_messages(
+        query="那家族成员这一部分呢？",
+        history=history,
+        session_id="product-follow-up-session",
+        job_id="job-product-follow-up",
+        decision=RouteDecision(
+            goal="general",
+            needs=[],
+            subject="继续解释产品业务",
+            interaction_kind="supplement",
+        ),
+        goal_ledger=None,
+    )
+
+    assert [message.content for message in messages[1:-1]] == [
+        item["content"] for item in history
+    ]
+
+
+def test_completed_subfamily_or_stage_list_is_not_full_analysis_completion():
+    assert not RouterRuntime._claims_gene_family_complete_at_readiness_boundary(
+        "亚家族分类与确认已经完成，接下来进入常规基因家族分析，包括 ppi。"
+    )
+    assert not RouterRuntime._claims_gene_family_complete_at_readiness_boundary(
+        "已完成的阶段包括 identify；后续还有常规基因家族分析。"
+    )
+    assert RouterRuntime._claims_gene_family_complete_at_readiness_boundary(
+        "亚家族确认完成，因此完整基因家族分析已经全部完成。"
+    )

tests/unit/test_persistent_conversation_full_history.py

@@ -0,0 +1,100 @@
+from gfa_core.mas.agent_state_repository import SQLiteAgentStateRepository
+from gfa_core.mas.agents.router_runtime import RouteDecision, RouterRuntime
+from gfa_core.mas.persistent_conversation import PersistentConversationStore
+
+
+def test_begin_turn_imports_and_returns_complete_valid_history(tmp_path):
+    repository = SQLiteAgentStateRepository(tmp_path / "agent-state.sqlite3")
+    repository.initialize()
+    store = PersistentConversationStore(repository)
+    client_history = [
+        {
+            "role": "user" if index % 2 == 0 else "assistant",
+            "content": f"历史消息-{index:02d}",
+        }
+        for index in range(14)
+    ]
+
+    turn = store.begin_turn(
+        session_id="full-history-session",
+        job_id="job-full-history",
+        query="当前问题",
+        client_history=client_history,
+    )
+
+    assert list(turn.history) == client_history
+    stored = repository.list_messages(turn.conversation_id)
+    assert [item.content for item in stored] == [
+        *(item["content"] for item in client_history),
+        "当前问题",
+    ]
+
+
+def test_begin_turn_returns_all_persisted_messages_not_only_latest_ten(tmp_path):
+    repository = SQLiteAgentStateRepository(tmp_path / "agent-state.sqlite3")
+    repository.initialize()
+    store = PersistentConversationStore(repository)
+    first_turn = store.begin_turn(
+        session_id="persisted-full-history-session",
+        job_id="job-full-history",
+        query="第一个问题",
+        client_history=[],
+    )
+    store.finish_turn(
+        first_turn,
+        assistant_message="第一个回答",
+        interaction_kind="independent",
+        action="reply",
+    )
+    for index in range(1, 7):
+        turn = store.begin_turn(
+            session_id="persisted-full-history-session",
+            job_id="job-full-history",
+            query=f"问题-{index}",
+            client_history=[],
+        )
+        store.finish_turn(
+            turn,
+            assistant_message=f"回答-{index}",
+            interaction_kind="supplement",
+            action="reply",
+        )
+
+    final_turn = store.begin_turn(
+        session_id="persisted-full-history-session",
+        job_id="job-full-history",
+        query="最终问题",
+        client_history=[],
+    )
+
+    assert len(final_turn.history) == 14
+    assert final_turn.history[0]["content"] == "第一个问题"
+    assert final_turn.history[-1]["content"] == "回答-6"
+
+
+def test_main_agent_prompt_receives_the_complete_history():
+    history = [
+        {
+            "role": "user" if index % 2 == 0 else "assistant",
+            "content": f"消息-{index}",
+        }
+        for index in range(12)
+    ]
+
+    messages = RouterRuntime._initial_messages(
+        query="当前问题",
+        history=history,
+        session_id="full-history-session",
+        job_id="job-full-history",
+        decision=RouteDecision(
+            goal="general",
+            needs=[],
+            subject="继续普通问答",
+            interaction_kind="supplement",
+        ),
+        goal_ledger=None,
+    )
+
+    assert [message.content for message in messages[1:-1]] == [
+        f"消息-{index}" for index in range(12)
+    ]

tests/unit/test_router_analysis_handoff_boundary.py

@@ -19,6 +19,7 @@ class _PassThroughTools:
         *,
         workflow_id: str | None = None,
         guidance: bool = False,
+        language: str = "zh",
     ) -> str:
         return query
 
@@ -30,6 +31,7 @@ class _ContextEnrichingTools:
         *,
         workflow_id: str | None = None,
         guidance: bool = False,
+        language: str = "zh",
     ) -> str:
         return (
             f"{query}\n已确认上下文：当前 job 的 subfamily_confirm 已成功完成，"
@@ -150,3 +152,23 @@ def test_new_analysis_decision_discards_main_agent_workflow_and_clarification()
     assert normalized.workflow_id is None
     assert normalized.clarification_question is None
     assert normalized.analysis_request_kind == "execution"
+
+
+def test_router_response_language_does_not_rewrite_business_query() -> None:
+    query = "执行接下来的分析"
+    messages = RouterRuntime._initial_messages(
+        query=query,
+        history=[],
+        session_id="english-router-session",
+        job_id="job-english-router",
+        decision=RouteDecision(
+            goal="general",
+            needs=[],
+            subject=query,
+        ),
+        goal_ledger=None,
+        language="en",
+    )
+
+    assert messages[-1].content == query
+    assert "Respond to the user in English." in messages[0].content

tests/unit/test_stage_requirements_context_efficiency.py

@@ -0,0 +1,107 @@
+from types import SimpleNamespace
+
+from gfa_core.mas import context_index, stage_requirements
+from gfa_core.mas.context_index import ProjectContextIndex
+from gfa_core.mas.stage_requirements import (
+    resolve_many_stage_requirements,
+    resolve_stage_requirements,
+)
+
+
+def test_stage_requirements_does_not_load_unconsumed_global_stage_status(
+    tmp_path,
+    monkeypatch,
+):
+    calls = []
+
+    def fake_get_stage_status(**kwargs):
+        calls.append(kwargs["stage"])
+        return SimpleNamespace(status="missing", manifest_path="", to_dict=lambda: {})
+
+    monkeypatch.setattr(context_index, "get_stage_status", fake_get_stage_status)
+
+    report = resolve_stage_requirements(
+        stage="identify",
+        prefix="Ft.SPL",
+        job_id="job-test",
+        project_root=tmp_path,
+    )
+
+    assert report.stage == "identify"
+    assert calls == []
+
+
+def _empty_context(tmp_path, job_id):
+    workspace = tmp_path / "workspace" / job_id
+    return ProjectContextIndex(
+        job_id=job_id,
+        workspace=str(workspace),
+        input_dir=str(tmp_path / "input" / job_id),
+        final_dir=str(workspace / "final"),
+        logs_dir=str(workspace / "logs"),
+    )
+
+
+def test_many_stage_requirements_reuses_context_without_mutating_shared_identity(
+    tmp_path,
+    monkeypatch,
+):
+    calls = []
+    base_context = _empty_context(tmp_path, "job-test")
+
+    def fake_build_project_context(**kwargs):
+        calls.append(kwargs["prefix"])
+        return base_context
+
+    monkeypatch.setattr(
+        stage_requirements,
+        "build_project_context",
+        fake_build_project_context,
+    )
+
+    reports = resolve_many_stage_requirements(
+        stages=["identify", "phylogeny"],
+        prefix="Ft.SPL",
+        job_id="job-test",
+        params_by_stage={
+            "identify": {"targetSpecies": ["Ft"]},
+            "phylogeny": {"targetSpecies": ["Os"]},
+        },
+        project_root=tmp_path,
+    )
+
+    assert calls == ["Ft.SPL"]
+    assert base_context.target_species == []
+    assert reports["identify"]["identity"]["target_species"] == ["Ft"]
+    assert reports["phylogeny"]["identity"]["target_species"] == ["Os"]
+
+
+def test_many_stage_requirements_caches_context_per_effective_prefix(
+    tmp_path,
+    monkeypatch,
+):
+    calls = []
+
+    def fake_build_project_context(**kwargs):
+        calls.append(kwargs["prefix"])
+        return _empty_context(tmp_path, kwargs["job_id"])
+
+    monkeypatch.setattr(
+        stage_requirements,
+        "build_project_context",
+        fake_build_project_context,
+    )
+
+    resolve_many_stage_requirements(
+        stages=["identify", "phylogeny", "motif"],
+        prefix="Ft.SPL",
+        job_id="job-test",
+        params_by_stage={
+            "identify": {"prefix": "Ft.SPL"},
+            "phylogeny": {"prefix": "Ft.SPL"},
+            "motif": {"prefix": "At.SPL"},
+        },
+        project_root=tmp_path,
+    )
+
+    assert calls == ["Ft.SPL", "At.SPL"]