{"avoid_when":"只要基准跑分（lm-eval）或提示词红队（promptfoo）","content_hash":"251b7825fb90e3fcccb8327ce1d43fe421901a2f4704a3bcf09188e550d0d5f6","domain":"ai-agents","evidence_urls":["https://github.com/truera/trulens"],"id":"trulens","language":"Python","license":"MIT","name":"TruLens","niche":"llm eval feedback","repo":"https://github.com/truera/trulens","source_updated_at":"2026-07-21T20:48:50Z","status":"active","summary":"LLM/RAG 应用评估与可观测反馈库，偏反馈回路。","tags":["eval","observability"],"use_when":"要把 RAG/Agent 质量评估接到应用反馈闭环","verification_status":"verified","verified_at":"2026-07-21T20:48:50Z"}