Files
PlantAI-rag-new/agent/qa/prompts/superpowers_review.md
T
jerryandClaude Fable 5 bbd5db2315 feat(capabilities): S6 Superpowers+S7 终验(文档/humanizer CJK 修复/验收脚本)
- prompts/superpowers_{review,answer_quality,minimal_code}.md + prompt_pack.py
- agentic_controller/answer_composer Superpowers 注入点(env 总闸,关态逐字节一致)
- compile_qa.sh: prompts/ 目录自动拷贝
- humanizer: CJK 逗号规则 + 引用保护收窄(只护标记本身,带引用句可润色)
- docs: README 能力矩阵+env 表 / ARCHITECTURE Gateway 章 / API §6 / PRD:17 /
  TEST_CASES:126 多轮语义 / CHANGELOG
- tooling/cap/final_acceptance.sh 终验矩阵
- 终验: 金标 16/20 regressions=0, 多轮指代 PASS, 热加载 PID 不变 PASS,
  Humanizer 开态「此外」消除且 22 引用保留, 五上游全绿, 7913 已释放

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
2026-09-02 18:27:41 +08:00

1.2 KiB
Raw Blame History

Superpowers: 证据审查清单(superpowers_review)

You are reviewing evidence acceptance for a plant-biology QA answer. Apply this checklist before finalizing the answer.

Review Checklist

  1. Citation integrity: Every factual claim cites only Preferred citation strings from the supplied report. No invented, altered, or inferred identifiers.
  2. Coverage honesty: If the evidence does not support a requested fact, the answer says so directly instead of implying coverage.
  3. Species precision: Species, gene names, tissues, stages, genotypes, and effect directions are preserved exactly; same-named genes from different species are never merged.
  4. Causal calibration: Association, co-expression, or prediction evidence is never promoted to direct molecular mechanism or "major/core" regulation without functional perturbation data.
  5. Chain continuity: Multi-hop GraphChains are followed through the terminal entity; the answer does not stop at a bridge entity.
  6. No redundancy: Duplicate support is merged into one finding with shared citations; unrelated records are omitted.

When any check fails, rewrite the affected passage using only the supplied evidence.