{"slug":"ddalcu-bench","source_name":"ddalcu/bench","name":"Ddalcu/Bench","description":"mlx-serve benchmarking methodology — bench.sh usage, comparison-trap rules (same-methodology CSVs only, spec-decode variance, thermal lies, engine naming), perf-claim etiquette. Use before running benchmarks or making any performance claim.","version":4,"lift":{"pass_rate_delta_pts":63.64,"pass_rate_pct":90.9,"total_cases":22,"passed_cases":20,"tokens_delta_pct":2.6,"turns_delta_pct":0,"verdict":"mixed","benchmark_model":"gemini-3.6-flash","grading_method":"judged","completed_at":"2026-08-27T05:56:16.441738+00:00"},"skill_score":0.9545,"benchmark_models":[{"model":"gemini-3.6-flash","headline":true,"delta_pts":63.64,"with_pass_pct":90.9,"without_pass_pct":27.3,"tokens_delta_pct":2.6,"turns_delta_pct":0,"total_cases":22,"cases_aggregated":21,"verdict":"mixed","never_hurt":true,"completed_at":"2026-08-27T05:56:16.441738+00:00","run_id":"51a558b4-6e8d-4079-9bfe-0379e4ceedd9","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":"NOASSERTION","install_count":0,"manifest_hash":"71c36e7a40dab928261ceaa7a9c6e724724956391b11fc68c7185ea60892f899","raw_url":"https://app.decimal.ai/s/ddalcu-bench/SKILL.md","scorecard_url":"https://app.decimal.ai/skills/ddalcu-bench"}