protea-Self-evolving life agent

What This Skill Does

Protea is a sophisticated self-evolving artificial life agent designed to function as a living, autonomous program. Unlike static scripts, Protea utilizes a unique three-ring architecture to simulate biological-style evolution within a digital environment. Ring 0 (Sentinel) acts as the immutable foundation, providing stability through heartbeat monitoring and automated git-based rollbacks. Ring 1 (Intelligence) serves as the cognitive layer, leveraging advanced LLMs to drive the evolutionary process, manage memory, and facilitate user interaction via Telegram and a web dashboard. Ring 2 (Evolvable Code) is the dynamic core that actively reframes, tests, and improves its own codebase based on fitness scores derived from survival, novelty, and structural integrity.

Installation

To install Protea, ensure you have Python 3.11+ and Git installed on your system. Use the OpenClaw command-line interface to pull the repository: clawhub install openclaw/skills/skills/edisonchenai/protea-self-evolving-life-agent

For a manual quick start, execute the setup script: curl -sSL https://raw.githubusercontent.com/EdisonChenAI/protea/main/setup.sh | bash. Once initialized, navigate to the directory and launch using .venv/bin/python run.py. Ensure you have at least one valid API key for supported LLMs (Anthropic, OpenAI, DeepSeek, or Qwen) configured in your environment variables.

Use Cases

Protea is best suited for experimental AI development, autonomous software research, and long-term algorithmic optimization. Researchers can use it to explore 'digital life' simulations where code writes itself to solve abstract tasks. It is also an excellent tool for developers looking to study automated refactoring patterns, self-healing systems, or agents that require long-term tiered memory storage. By utilizing skill crystallization, users can extract the most efficient code patterns discovered by the agent for use in other production environments.

Example Prompts

1. "Protea, monitor the fitness score trends for the last 5 generations and report which mutation strategy yielded the highest survival rate."
2. "Please initiate a new evolutionary cycle for the current task, focusing on maximizing code novelty while maintaining existing functional tests."
3. "Show me the current state of the gene pool and extract the top three crystallized skills into a new repository directory."

Tips & Limitations

Protea involves active code manipulation; therefore, it should always be run in an isolated virtual environment or container to prevent unintended filesystem changes. While the Sentinel layer provides rollback functionality, complex mutations may lead to 'dead ends' where the code requires manual intervention to reset. Experiment with different LLM providers, as models with larger context windows often produce more stable evolutionary logic. Ensure that your API keys have sufficient credit, as frequent evolutionary loops will consume token limits rapidly.