{"slug":"affaan-m-benchmark","source_name":"affaan-m/benchmark","name":"Affaan M/Benchmark","description":"このスキルを使用して、パフォーマンスベースラインを測定し、PR前後の回帰を検出し、スタック代替案を比較します。","version":4,"lift":{"pass_rate_delta_pts":45.45,"pass_rate_pct":90.9,"total_cases":22,"passed_cases":20,"tokens_delta_pct":-1.3,"turns_delta_pct":0,"verdict":"mixed","benchmark_model":"gemini-3.6-flash","grading_method":"judged","completed_at":"2026-08-12T12:10:04.102094+00:00"},"skill_score":0.9091,"benchmark_models":[{"model":"gemini-3.6-flash","headline":true,"delta_pts":45.45,"with_pass_pct":90.9,"without_pass_pct":45.5,"tokens_delta_pct":-1.3,"turns_delta_pct":0,"total_cases":22,"cases_aggregated":21,"verdict":"mixed","never_hurt":false,"completed_at":"2026-08-12T12:10:04.102094+00:00","run_id":"ed2a1825-14d7-417b-b2f2-18ebb764442f","version_number":4,"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":"da7ea8f4da0d53296a04b18656a86f541e587691cff9abab554d382650cd28de","raw_url":"https://app.decimal.ai/s/affaan-m-benchmark/SKILL.md","scorecard_url":"https://app.decimal.ai/skills/affaan-m-benchmark"}