{"slug":"majiayu000-agent-evals","source_name":"majiayu000/agent-evals","name":"Majiayu000/Agent Evals","description":"Design and implement evaluation frameworks for AI agents. Use when testing agent reasoning quality, building graders, doing error analysis, or establishing regression protection. Framework-agnostic concepts that apply to any SDK.","version":1,"lift":{"pass_rate_delta_pts":26.09,"pass_rate_pct":78.3,"total_cases":23,"passed_cases":18,"tokens_delta_pct":94.1,"turns_delta_pct":0,"verdict":"mixed","benchmark_model":"gemini-3.6-flash","grading_method":"judged","completed_at":"2026-09-03T05:34:23.584124+00:00"},"skill_score":0.881866037735849,"benchmark_models":[{"model":"gemini-3.6-flash","headline":true,"delta_pts":26.09,"with_pass_pct":78.3,"without_pass_pct":52.2,"tokens_delta_pct":94.1,"turns_delta_pct":0,"total_cases":23,"cases_aggregated":23,"verdict":"mixed","never_hurt":true,"completed_at":"2026-09-03T05:34:23.584124+00:00","run_id":"123033d1-2658-4f2d-8338-e5d8d3d1ff5a","latest_attempt":{"id":"123033d1-2658-4f2d-8338-e5d8d3d1ff5a","runner_model":"gemini-3.6-flash","overall_verdict":"mixed","completed_at":"2026-09-03T05:34:23.584124+00:00","gate":null},"version_number":1,"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,"total_uses":76,"total_users":3,"manifest_hash":"9aa18b3879ec99415cedd87656c71f68a0d5345f474f08c38270e153468582db","install_name":"majiayu000-agent-evals","version_id":"34398bf8-a8de-4a03-b5eb-1d5a9e2db97b","body_sha256":"9aa18b3879ec99415cedd87656c71f68a0d5345f474f08c38270e153468582db","render_hash":"a06e8ffd2eb45aec94512a373550d804b541946419a24687f08fb2c85b34e470","raw_url":"https://app.decimal.ai/s/majiayu000-agent-evals/SKILL.md","scorecard_url":"https://app.decimal.ai/skills/majiayu000-agent-evals"}