SkillClaude CodeCursorClineOpenClawWindsurfVS CodeChatGPT
proactive-self-improving-agent
Proactive Self-Improving Agent skill for OpenClaw - combines context protection (WAL/Buffer) with structured learning and safe evolution
Features
- ✓7 trigger conditions for proactive learning
- ✓Structured record keeping for LEARNINGS, ERRORS, FEATURE_REQUESTS
- ✓Experience evolution with promotion and recursive detection
- ✓Safety guards with ADL and VFM for value prioritization
- ✓Operational logs in JSONL format for machine readability
- ✓Guided by principles of perseverance and safety reinforcement
- ✓Explore more at [GitHub](https://github.com/yanhongxi-openclaw/proactive-self-improving-agent)
Installation Guide
This is a curated collection of tools. To install, browse the GitHub repository for individual tool installations and follow their specific instructions.