{"slug":"devtu-benchmark-harness","source_name":"devtu-benchmark-harness","name":"Devtu Benchmark Harness","description":"Continuous improvement system for ToolUniverse tools, skills, and plugin. Run benchmarks, diagnose failures, route fixes to devtu skills, retest. Use after skill optimization, tool additions, or as regression check.","version":1,"lift":{"pass_rate_delta_pts":68,"pass_rate_pct":72,"total_cases":25,"passed_cases":18,"tokens_delta_pct":106.6,"turns_delta_pct":0,"verdict":"mixed","benchmark_model":"gemini-3.6-flash","grading_method":"judged","completed_at":"2026-07-27T21:43:46.929829+00:00"},"skill_score":0.7199999999999999,"benchmark_models":[{"model":"gemini-3.6-flash","headline":true,"delta_pts":68,"with_pass_pct":72,"without_pass_pct":4,"tokens_delta_pct":106.6,"turns_delta_pct":0,"total_cases":25,"cases_aggregated":22,"verdict":"mixed","never_hurt":true,"completed_at":"2026-07-27T21:43:46.929829+00:00","run_id":"109f6117-494f-470a-ad92-a1d8324c66e0","latest_attempt":{"id":"109f6117-494f-470a-ad92-a1d8324c66e0","runner_model":"gemini-3.6-flash","overall_verdict":"mixed","completed_at":"2026-07-27T21:43:46.929829+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":"Apache License 2.0","install_count":0,"total_uses":11,"total_users":1,"manifest_hash":"7b38fc869da8ced13c545f27bfffdbfa5b7477c505869232ed2348be971e8566","install_name":"devtu-benchmark-harness","version_id":"8dab4856-a382-4a46-a870-0ff2b75f0391","body_sha256":"7b38fc869da8ced13c545f27bfffdbfa5b7477c505869232ed2348be971e8566","render_hash":"b7033c64b1d92f0d1e73c9dd2e8390662c58440c9ce8ddd76edb3088b41b8ff6","raw_url":"https://app.decimal.ai/s/devtu-benchmark-harness/SKILL.md","scorecard_url":"https://app.decimal.ai/skills/devtu-benchmark-harness"}