{"slug":"svgreg-sg-llm-polish","source_name":"svgreg/sg-llm-polish","name":"Svgreg/Sg LLM Polish","description":"Improve skill-guard's optional LLM (T3 semantic) analysis prompt — fold in newly-discovered cases, harden it against hijack by scanned content, and check efficiency. Guarded stub while the LLM engine is unimplemented (M5); it self-checks and no-ops, accumulating prompt-design notes instead. Use when asked to polish the LLM prompt or improve semantic analysis.","version":1,"lift":{"pass_rate_delta_pts":40.91,"pass_rate_pct":68.2,"total_cases":22,"passed_cases":15,"tokens_delta_pct":-7.3,"turns_delta_pct":0,"verdict":"mixed","benchmark_model":"gemini-3.6-flash","grading_method":"judged","completed_at":"2026-08-04T02:08:56.482092+00:00"},"skill_score":0.6818,"benchmark_models":[{"model":"gemini-3.6-flash","headline":true,"delta_pts":40.91,"with_pass_pct":68.2,"without_pass_pct":27.3,"tokens_delta_pct":-7.3,"turns_delta_pct":0,"total_cases":22,"cases_aggregated":17,"verdict":"mixed","never_hurt":false,"completed_at":"2026-08-04T02:08:56.482092+00:00","run_id":"036077c1-4086-4d57-8a8c-bb6edc4bf16a","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":null,"install_count":0,"manifest_hash":"6060f0240eec56f8605be98bfc0a0ad81e5a5740444631fedbc5fec4b7d7bfd1","raw_url":"https://app.decimal.ai/s/svgreg-sg-llm-polish/SKILL.md","scorecard_url":"https://app.decimal.ai/skills/svgreg-sg-llm-polish"}