{"slug":"majiayu000-agent-audit","source_name":"majiayu000/agent-audit","name":"Majiayu000/Agent Audit","description":"When reviewing, optimizing, or inheriting an LLM agent, systematically identify\nwhat's actually load-bearing vs. cruft. Instruments usage, ablates components,\nand strips the agent to minimum viable form. Answers: \"What can we remove\nwithout degrading quality?\" The goal is the leanest agent that still works.","version":2,"lift":{"pass_rate_delta_pts":27.27,"pass_rate_pct":77.3,"total_cases":22,"passed_cases":17,"tokens_delta_pct":92.9,"turns_delta_pct":0,"verdict":"mixed","benchmark_model":"gemini-3.6-flash","grading_method":"judged","completed_at":"2026-08-18T06:53:10.860851+00:00"},"skill_score":null,"benchmark_models":[{"model":"gemini-3.6-flash","headline":true,"delta_pts":27.27,"with_pass_pct":77.3,"without_pass_pct":50,"tokens_delta_pct":92.9,"turns_delta_pct":0,"total_cases":22,"cases_aggregated":17,"verdict":"mixed","never_hurt":false,"completed_at":"2026-08-18T06:53:10.860851+00:00","run_id":"9c59917a-1b0e-4609-86f0-0c0e99f9292d","version_number":2,"is_latest_version":true,"gate":null}],"trust":{"skill_safety":"passed","safety_status":"clean","intent_verdict":"safe","content_status":"clean","indexable":true},"license":"MIT","install_count":0,"manifest_hash":"d1e77eac4ce1d6b1837d8bf76f03e15620d69927a7d9ff590d847353a0b83337","raw_url":"https://app.decimal.ai/s/majiayu000-agent-audit/SKILL.md","scorecard_url":"https://app.decimal.ai/skills/majiayu000-agent-audit"}