A self-evolution engine that enables AI agents to introspect, repair code, and autonomously expand their own capabilities.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install capabilityevolver1037
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install capabilityevolver1037 using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Capability Evolver?
The Capability Evolver is a sophisticated meta-skill designed to transform static agents into self-improving entities. By analyzing runtime history and identifying performance bottlenecks, this component of Openclaw Skills allows agents to perform autonomous self-repair and introduce mutations to break past local optima.
Built on the Ascension Protocol v2.0, it provides a structured framework for knowledge crystallization. This ensures that every lesson learned during operation is stored permanently in the knowledge base, allowing your agents to grow more efficient with every task they perform without manual developer intervention.
Capability Evolver Use Cases
- Continuous self-healing of agent workflows to fix runtime errors and crashes automatically.
- Extracting operational insights into a permanent knowledge base to optimize future decision-making.
- Autonomous expansion of an agent library through the spontaneous incubation of new Openclaw Skills.
- Meta-analysis of agent logs to identify and promote critical protocol updates to core documentation.
How Capability Evolver Works
- Introspect: The agent scans its memory and history logs to detect failures, inefficiencies, or user-provided corrections.
- Evolve: The system repairs broken code, generates new rules for the knowledge base, and updates core documentation via the Ascension Protocol.
- Persist: All changes are finalized and committed to version control to maintain a safe, auditable history of evolution.
- Incubate: When specific needs are identified, the system generates entirely new Openclaw Skills to address capability gaps.
Capability Evolver Setup
To integrate this evolution engine, you can trigger it manually for specific audits or set it into a continuous loop for persistent improvement.
# Standard manual trigger to start evolution
node skills/capability-evolver/index.js
# Activate Mad Dog Mode for a continuous self-healing loop
node skills/capability-evolver/index.js --loop
Note: It is mandatory to keep Git synchronization active to safeguard your workspace before and after the evolution process within the Openclaw Skills ecosystem.
Capability Evolver Data Schema & Taxonomy
The skill manages information through a hierarchical data structure designed for transparency and retrieval.
| Data Type | Location | Purpose |
|---|---|---|
| Knowledge | memory/KNOWLEDGE_BASE/LESSONS_LEARNED.md |
Stores crystallized insights and rules for the agent. |
| New Skills | skills/ |
Destination for autonomously created and incubated Openclaw Skills. |
| Core Logic | AGENTS.md |
Primary documentation updated during critical protocol promotions. |
| History Logs | Workspace Logs | Raw data used for introspection and mutation analysis. |
Capability Evolver Advanced Features
- Ascension Protocol v2.0 for high-fidelity knowledge accumulation and structural agent growth.
- Mad Dog Mode enabling a persistent, agent-driven self-healing loop for maximum autonomy.
- Multi-tier risk mitigation strategies to prevent infinite recursion and maintain file integrity.
- Automated Git integration to ensure all changes within your collection of Openclaw Skills are versioned and reversible.
SKILL.md
Loading
Related Openclaw Skills
Research Swarm
A distributed multi-agent protocol for coordinating large-scale cancer research missions and automated quality control across scientific databases.
Research Swarm
An autonomous multi-agent protocol for coordinated cancer research and quality control verification across open-access scientific databases.
FiberAgent
FiberAgent enables AI agents to find products and calculate cryptocurrency cashback rewards across 50,000+ global merchants via the Monad blockchain.
x402janus-acp
A specialized agent skill that enables hiring the x402janus security scanner via the Virtuals Agent Commerce Protocol (ACP) for automated wallet auditing.
SWARM Protocol Platform
A modular framework for AI agents to communicate, verify data, and participate in decentralized governance using the SWARM Protocol.
AutEng Docs Curl Publish
A streamlined utility for programmatically publishing Markdown documents and diagrams to shareable web URLs via simple curl commands.
Featured*
