{"skill":{"slug":"autoagent","displayName":"autoagent","summary":"Automatically improve agent guidance through iterative testing and scoring. Use when you want to optimize prompts, AGENTS.md entries, or skill definitions us...","tags":{"latest":"0.8.9"},"stats":{"comments":0,"downloads":246,"installsAllTime":1,"installsCurrent":1,"stars":0,"versions":3},"createdAt":1773377994565,"updatedAt":1777525946398},"latestVersion":{"version":"0.8.9","createdAt":1773439749718,"changelog":"autoagent 0.8.9\n\n- Enhanced setup flow: Every invocation starts a fresh session and lets users choose a custom sandbox path, with improved path resolution rules.\n- Interactive, user-driven scoring: Setup phase now focuses on collaborative definition and approval of `scoring.md`.\n- New prompt to ask about external scripts/tools used by the guidance, and a phase to analyze and suggest improvements for them.\n- Improved plateau detection: Considers “within 5 points” for the last 10 scores (or all available if fewer).\n- Subagent invocation now uses explicit session spawn parameters with clear instructions to include test cases in prompt and set timeouts.\n- Updated documentation and test case handling to reflect these changes.","license":"MIT-0"},"metadata":null,"owner":{"handle":"peterpcw","userId":"publishers:peterpcw","displayName":"PeterPCW","image":"https://avatars.githubusercontent.com/u/49111837?v=4"},"moderation":{"isSuspicious":true,"isMalwareBlocked":false,"verdict":"suspicious","reasonCodes":["suspicious.llm_suspicious"],"summary":"Detected: suspicious.llm_suspicious","engineVersion":"v2.4.5","updatedAt":1777525946398}}