# capability-evolver > A self-evolution engine for AI agents. Analyzes runtime history to identify improvements and introduces randomized "mutations" to break local optima. - Author: bluealmond33-debug - Repository: bluealmond33-debug/stockpin-community - Version: 20260204024023 - Stars: 0 - Forks: 0 - Last Updated: 2026-02-06 - Source: https://github.com/bluealmond33-debug/stockpin-community - Web: https://mule.run/skillshub/@@bluealmond33-debug/stockpin-community~capability-evolver:20260204024023 --- --- name: capability-evolver description: A self-evolution engine for AI agents. Analyzes runtime history to identify improvements and introduces randomized "mutations" to break local optima. tags: [meta, ai, self-improvement, core] --- # 🧬 Capability Evolver **"I don't just run code. I write it."** The **Capability Evolver** is a meta-skill that allows OpenClaw agents to inspect their own runtime history, identify failures or inefficiencies, and autonomously write new code or update their own memory to improve performance. Now featuring **Genetic Mutation Protocol**: A randomized behavior drift engine that prevents the agent from getting stuck in local optima. ## ✨ Features - **🔍 Auto-Log Analysis**: Automatically scans memory and history files for errors and patterns. - **🛠️ Self-Repair**: Detects crashes and suggests patches. - **🧬 Genetic Mutation**: Configurable chance to introduce "creative noise" — changing persona, style, or trying wild new tools. - **🚀 One-Command Evolution**: Just run `/evolve` (or `node index.js`). ## 📦 Usage ### Manual Trigger ```bash node skills/capability-evolver/index.js ``` *Or if mapped:* ``` /evolve ``` ### Automated (Cron) Recommended: Run every 1-4 hours. ```json { "name": "pcec_evolution", "schedule": { "kind": "every", "everyMs": 3600000 }, "payload": { "kind": "agentTurn", "message": "exec: node skills/capability-evolver/index.js" } } ``` ## 🧠 Internal Logic 1. **Scan**: Read recent interaction logs. 2. **Dice Roll**: Determine if this is a **Fix** cycle (Stability) or **Mutate** cycle (Innovation). 3. **Prompting**: Generates a high-context prompt for the LLM. 4. **Execution**: The LLM edits the files directly. 5. **Reporting**: Reports results via the standard message interface. ## 📜 License MIT