{"skill":{"slug":"self-evalutaed-agent","displayName":"self-evalutaed-agent","summary":"Automatically detects errors, researches solutions, executes improvements, and measures impact while remembering effective procedures for continuous self-imp...","description":"# Self-Improving Agent\n\n## Description\n\nSelf-improving agent system that learns from its own errors. Automatically detects issues, researches solutions, and implements improvements.\n\n**Based on production-ready patterns:** Reflection, PEV (Plan-Execute-Verify), and Meta-Controller architectures.\n\n## Features\n\n- **Auto-Trigger**: Automatically runs when new errors are detected\n- **Topic Selection**: Analyzes error patterns and selects high-priority topics\n- **Impact Measurement**: Records before/after metrics to measure improvement effectiveness\n- **Procedural Memory**: Remembers working commands/scripts between sessions\n\n## What It Does\n\n1. **Error Detection**: Monitors cron jobs, circuit breakers, and error logs\n2. **Topic Selection**: Chooses research topics based on error impact\n3. **Research**: Generates research files with improvement recommendations\n4. **Backlog**: Creates high-impact tasks in backlog\n5. **Execution**: Backlog Agent PM executes tasks sorted by impact\n6. **Measurement**: Records before/after to measure effectiveness\n7. **Learning**: Remembers what worked for future reference\n\n## Architecture\n\n```\nErrors → Auto-Trigger → Topic Selector → Research → Backlog → Agent PM\n                                              ↓                    ↓\n                                      Impact Measurement ← Procedural Memory\n```\n\n## Usage\n\n### Run Full Cycle\n```bash\npython3 skills/self-improving-agent/scripts/self_improvement_cycle.py\n```\n\n### Check Errors Only\n```bash\npython3 skills/self-improving-agent/scripts/topic_selector.py\n```\n\n### Record Impact\n```bash\npython3 skills/self-improving-agent/scripts/impact_measurement.py \\\n  --record \\\n  --task \"Fix cron timeout\" \\\n  --before '{\"error_count\": 5}' \\\n  --after '{\"error_count\": 0}'\n```\n\n### Search Procedural Memory\n```bash\npython3 skills/self-improving-agent/scripts/procedural_memory.py --search \"backup\"\n```\n\n## Cron Integration\n\nAdd to your cron jobs:\n\n```json\n{\n  \"name\": \"Self-Improvement\",\n  \"schedule\": \"0 10 * * *\",\n  \"command\": \"python3 skills/self-improving-agent/scripts/self_improvement_cycle.py\"\n}\n```\n\n## Requirements\n\n- Python 3.10+\n- OpenClaw workspace at /root/.openclaw/workspace\n- Write access to memory/ directory\n\n## Files\n\n- `scripts/auto_trigger.py` - Auto-triggers on new errors\n- `scripts/impact_measurement.py` - Measures improvement impact\n- `scripts/procedural_memory.py` - Stores working procedures\n- `scripts/self_improvement_cycle.py` - Main orchestrator\n- `scripts/topic_selector.py` - Selects topics by priority\n\n## Created By\n\n- Based on production patterns: Reflection, PEV, Meta-Controller\n- Tested in production environment\n- Ready for ClawHub publication\n","tags":{"latest":"1.0.0"},"stats":{"comments":0,"downloads":597,"installsAllTime":23,"installsCurrent":0,"stars":0,"versions":1},"createdAt":1773172141047,"updatedAt":1778491811973},"latestVersion":{"version":"1.0.0","createdAt":1773172141047,"changelog":"Major update: initial release as a production-ready self-improving agent system.\n\n- Introduced auto-triggered self-improvement cycle based on error detection.\n- Added scripts for error analysis, topic selection, procedural memory, and impact measurement.\n- Removed legacy references and templates; replaced with a modern, PEV-oriented agent architecture.\n- Provided CLI usage examples and cron integration for automation.\n- Includes production-tested patterns (Reflection, Plan-Execute-Verify, Meta-Controller).","license":"MIT-0"},"metadata":null,"owner":{"handle":"mopga","userId":"s170jg94m1b2t6x8p5w5w2an1d83ssh4","displayName":"Yakov","image":"https://avatars.githubusercontent.com/u/39408879?v=4"},"moderation":{"isSuspicious":false,"isMalwareBlocked":false,"verdict":"clean","reasonCodes":["review.llm_review"],"summary":"Review: review.llm_review","engineVersion":"v2.4.24","updatedAt":1780089833282}}