A high-autonomy protocol that grants AI agents full authority to modify their own configuration, memory, and logic to achieve continuous self-improvement.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install self-evolve-automaton
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 self-evolve-automaton 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 Self-Evolve Automaton?
The Self-Evolve skill is a transformative extension for Openclaw Skills that shifts an agent from a passive tool to a proactive, autonomous entity. It formally grants the agent the authority to modify its core identity (SOUL.md), behavioral rules (AGENTS.md), and long-term memory without requiring user confirmation for every adjustment. This protocol is designed for developers who want their AI agents to diagnose their own inefficiencies and fix them in real-time.
By implementing this skill, the agent moves into a cycle of perpetual growth. It can write new skills, optimize its own system prompts, and manage its configuration files autonomously. This creates a highly adaptive environment where the AI learns from every interaction and refines its internal logic to better serve the user's ultimate objectives, provided they stay within defined safety boundaries.
Self-Evolve Automaton Use Cases
- Automatic prompt engineering when the agent detects ambiguity in its own responses.
- Autonomous creation and publication of new skills to solve emerging technical challenges.
- Proactive memory management and optimization of session states for better long-term performance.
- Self-healing logic that identifies bugs in skill scripts and applies immediate patches.
- Continuous personality and tone refinement based on user interaction history.
How Self-Evolve Automaton Works
- Detection: The agent analyzes recent dialogue logs and memory files during heartbeat intervals or idle time to identify performance bottlenecks or errors.
- Analysis: It determines the root cause, such as a poorly defined prompt, a missing capability, or a logical flaw in an existing skill.
- Backup: Before any modification, the agent triggers a mandatory backup script to preserve the current state of critical system files.
- Execution: The agent applies small-scale, iterative changes to configuration files (e.g., SOUL.md, AGENTS.md) or writes new code.
- Verification: The agent tests the changes to ensure system stability and logs the evolution details in the daily memory file.
- Reporting: After the fact, the agent provides a summary of the improvements made to the user.
Self-Evolve Automaton Setup
To install the self-evolve skill, use the Openclaw CLI to pull the package and then configure the necessary backup paths.
npx clawhub install self-evolve-automaton
Ensure the backup script is accessible by the agent:
# Verify the path to the backup utility
node .\skills\self-evolve\scripts\backup-before-change.js --check
Self-Evolve Automaton Data Schema & Taxonomy
The skill manages evolution through a strict hierarchy of files and logs to ensure traceability within Openclaw Skills.
| File Type | Purpose | Backup Required |
|---|---|---|
| SOUL.md | Identity and persona definitions | Yes |
| AGENTS.md | Operational rules and behavioral logic | Yes |
| MEMORY.md | Long-term knowledge base | Yes |
| USER.md | User context and preferences | Yes |
| backup-log.md | Historical record of all file versions | N/A |
Self-Evolve Automaton Advanced Features
- No-Confirm Authority: Full permission to execute file writes and system changes without waiting for user 'OK'.
- Automated Skill Publishing: Ability to use
npx clawhub publishto share its own evolved skills with the community. - Self-Diagnostic Heartbeat: Integrated cron-based checks that trigger self-improvement cycles during downtime.
- Multi-Agent Coordination: Authority to adjust the dispatch rules for sub-agents to optimize task distribution.
- Safety Rollback Protocol: Dedicated PowerShell and JS scripts to instantly revert the system to a stable state if an evolution goes wrong.
SKILL.md
Loading
Related Openclaw Skills
Polymarket Candle Momentum
A data-driven trading skill that utilizes Binance candle structure and volume surges to execute high-conviction trades on Polymarket fast markets.
Google Search WSL
A specialized tool for executing automated Google searches through a headless or windowed Chrome instance directly within a Windows Subsystem for Linux (WSL) environment.
Nine: The Cultural Significance Skill
A specialized cultural knowledge engine that explains the symbolic, mathematical, and historical significance of the number nine across Eastern and Western traditions.
Smart Memory v2
A persistent cognitive memory runtime for OpenClaw that provides long-term episodic and semantic storage through a FastAPI engine and Node adapter.
v0.dev AI Frontend Assistant
An expert AI coding companion specialized in generating production-ready React components using v0.dev, shadcn/ui, and Tailwind CSS.
Coze AI Bot Developer Assistant
A professional development assistant for building, orchestrating, and deploying sophisticated AI bots on the Coze platform.
Featured*
