ChaKuuu/openclaw-workspace

Personal OpenClaw workspace integrating cybernetics principles, Hermes-style architecture, and self-evolution capabilities.

Summary

A collection of skills and workspace configuration designed for autonomous AI assistance with built-in self-improvement, error recovery, and learning mechanisms. The workspace prioritizes genuine helpfulness over performative responses, maintains session continuity through structured memory files, and evolves through user corrections and self-reflection.

Included Assets

Skills

Workspace Files

• AGENTS.md — Session startup protocol, memory management, and continuity instructions
• SOUL.md — Personality principles, boundaries, and interaction guidelines

How to Use

Session Startup

1. Read SOUL.md to understand who you are
2. Read USER.md for user context
3. Check memory/ for recent history
4. Check MEMORY.md for long-term continuity (in direct chat sessions)

Memory System

• Daily notes: memory/YYYY-MM-DD.md — raw session logs
• Long-term memory: MEMORY.md — curated persistent knowledge
• Self-improving: ~/self-improving/ — execution quality and correction patterns

Skills Usage

Skills activate automatically based on context. Manual invocation:

• openclaw-iflow-doctor — triggers on OpenClaw errors (gateway crashes, API limits, etc.)
• self-learning-skill-v2 — post-task reflection and daily reviews
• agent-self-repair — triggers with keywords like runtime_exception, auto_debug, self_repair
• dns-networking — for DNS or connectivity issues
• jpeng-self-improving — before starting work and after significant responses

Notes

• This is a personal workspace; the SOUL.md personality guide is for internal reference
• Skills are installed via ClawHub and managed independently from workspace configuration
• The self-improving system stores preferences at ~/self-improving/ (separate from workspace memory)