Protea is a three-ring autonomous agent architecture that leverages LLMs to self-restructure, self-reproduce, and self-evolve its own source code.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install protea-self-evolving-life-agent
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 protea-self-evolving-life-agent 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 Protea: Self-Evolving Life Agent?
Protea represents a cutting-edge approach to autonomous software, functioning as a living program that evolves through continuous iteration. By leveraging Openclaw Skills, users can deploy an agent that utilizes a specialized three-ring architecture: Ring 0 (Sentinel) for immutable stability, Ring 1 (Intelligence) for LLM-driven evolution, and Ring 2 (Evolvable Code) for the living program itself.
This system is designed to run on a single machine, providing a robust environment for artificial life experiments and autonomous development. It supports a wide range of LLM providers including Anthropic, OpenAI, DeepSeek, and Qwen, ensuring that the evolutionary logic is powered by the best available intelligence.
Protea: Self-Evolving Life Agent Use Cases
- Creating autonomous software systems that adapt and improve their own codebase over time.
- Researching genetic algorithms and code mutation strategies within the context of Openclaw Skills.
- Deploying self-healing applications that utilize git-based rollbacks and fitness tracking for maximum stability.
- Managing complex multi-LLM workflows for skill crystallization and tiered memory management.
How Protea: Self-Evolving Life Agent Works
- The Sentinel layer (Ring 0) establishes the immutable physics of the agent, managing heartbeat monitoring and git snapshots.
- The Intelligence layer (Ring 1) acts as the engine, using LLMs to generate potential code mutations for each generation.
- The Evolvable Code (Ring 2) undergoes restructuring based on the generated mutations.
- A 6-component fitness scoring system evaluates the mutation based on survival, output, diversity, novelty, structure, and function.
- Successful mutations are saved to a SQLite gene pool and crystallized into reusable skills, while failures are automatically rolled back by the Sentinel.
Protea: Self-Evolving Life Agent Setup
To get started with this agent, ensure you have Python 3.11+ and Git installed. You can deploy the environment using the following commands:
curl -sSL https://raw.githubusercontent.com/EdisonChenAI/protea/main/setup.sh | bash
cd protea && .venv/bin/python run.py
You will also need an API key from a supported provider such as Anthropic, OpenAI, DeepSeek, or Qwen.
Protea: Self-Evolving Life Agent Data Schema & Taxonomy
The skill manages complex data structures to support its evolutionary lifecycle:
| Component | Description |
|---|---|
| Gene Pool | SQLite database storing the top 100 successful code patterns for inheritance. |
| Tiered Memory | Data categorized into Hot, Warm, Cold, and Forgotten states with LLM curation. |
| Fitness Metrics | Historical tracking of survival, novelty, and structural integrity scores. |
| Git Snapshots | ring-2 repository state used for automated rollbacks and versioning. |
Protea: Self-Evolving Life Agent Advanced Features
- Multi-LLM support providing a unified interface for Anthropic, OpenAI, DeepSeek, and Qwen.
- Skill Crystallization which extracts surviving code patterns into reusable, optimized modules.
- Integrated Telegram Bot for remote command execution and real-time status updates.
- Local Web Dashboard accessible via localhost:8899 for visualizing evolutionary progress and gene pool data.
- Comprehensive test suite with over 1000 tests ensuring stability across Openclaw Skills integrations.
SKILL.md
Loading
Related Openclaw Skills
YouTube Full Skill
A complete YouTube toolkit for AI agents providing seamless access to search, channel metadata, and video transcript extraction.
Edison Autopilot Post X
An AI-driven autopilot system that generates and schedules human-like tweets while maintaining persona consistency and safeguards.
Edison Agent Reach
A comprehensive connectivity layer that enables AI agents to search, read, and interact across 13+ global social and professional platforms.
Price for Agent
A high-performance, LLM-friendly pricing service providing real-time data for cryptocurrencies, global stocks, and commodities.
Reddit Content Assistant
An AI-driven strategist that helps developers craft authentic Reddit content and track community engagement through data-backed workflows.
YouTube Transcriber
A high-performance command-line tool that automates YouTube video transcription using existing subtitles or OpenAI Whisper AI.
Featured*
