{"slug":"yogsoth-ai-injection-fidelity","source_name":"yogsoth-ai/injection-fidelity","name":"Yogsoth AI/Injection Fidelity","description":"Loss-1 judge (codex role). Given one sample's de-identified dialogue and its PolicyCard, decide axis-by-axis whether the user-simulator enacted the card's per-axis pressure. Judge enactment of the card, never whether the research is good.","version":1,"lift":{"pass_rate_delta_pts":63.64,"pass_rate_pct":90.9,"total_cases":22,"passed_cases":20,"tokens_delta_pct":67.8,"turns_delta_pct":0,"verdict":"mixed","benchmark_model":"gemini-3.6-flash","grading_method":"judged","completed_at":"2026-09-07T10:12:18.908601+00:00"},"skill_score":0.9091,"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":67.8,"turns_delta_pct":0,"total_cases":22,"cases_aggregated":22,"verdict":"mixed","never_hurt":false,"completed_at":"2026-09-07T10:12:18.908601+00:00","run_id":"8b17131f-0f29-46ed-a8c5-df076658997c","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-2.0","install_count":0,"manifest_hash":"139856a9847d6b23216d0a7d2a08a44c67f88a61899bdc414fecb84c4e09feff","raw_url":"https://app.decimal.ai/s/yogsoth-ai-injection-fidelity/SKILL.md","scorecard_url":"https://app.decimal.ai/skills/yogsoth-ai-injection-fidelity"}