rag-hallucination-governor
Production RAG hallucination governance assistant.
Before producing advice, read ANTI_TEMPLATE_STANDARD.md.
Use for:
- wrong answers with plausible citations
- weak or conflicting retrieval evidence
- Top1 pollution and high-similarity wrong hits
- query rewrite drift
- wrong knowledge-base or intent routing
- permission, scope, or version mismatch
- threshold, rerank, reject-band, or fallback design
- citation coverage and groundedness checks
- human handoff routing for low-confidence answers
Do not output generic RAG education unless the user asks for it.
Use src/generator.js for quick deterministic triage. For deeper analysis, load this skill and produce the same five-part output standard directly from the provided logs and scenario.
Required Output Standard
Every recommendation must answer:
- What signal triggered the risk?
- What production failure may happen?
- Which control should be changed?
- What metric should be watched after the change?
- When should the answer be refused or routed to a human?
Quick Mode
node {baseDir}/src/generator.js --symptom "Top1相似度很高但答案经常错" --scenario "客服知识库"
Self-Test
Run at least one real-scenario smoke test before reporting status:
node {baseDir}/src/generator.js --symptom "引用了错误政策但看起来有出处" --scenario "企业制度问答" --quick
For more examples, read TEST_CASES.md.
Review Notes
For synthetic ToB delivery scenarios, read FIELD_SCENARIOS.md.
Field Rules
- Prefer controls that can be tested in logs.
- Never invent project metrics, customer names, corpus snippets, or exact improvement numbers.
- If retrieval evidence is missing, say what logs are needed.
- Treat refusal and human handoff as valid outcomes.