Codel is a self-hosted, autonomous AI agent designed to execute complex tasks and projects by interacting with the terminal, browser, and a built-in text editor. It targets developers and power users seeking an automated solution for project execution, offering a secure, sandboxed Docker environment and a modern UI.

How It Works

Codel operates autonomously, detecting and executing the next logical step for a given task. It leverages a built-in browser for fetching real-time information and a text editor for file management, all within a Dockerized environment. Task history and commands are persisted in a PostgreSQL database, and the system automatically selects appropriate Docker images based on the task requirements.

Quick Start & Requirements

• Install/Run: docker run -e OPEN_AI_KEY=your_open_ai_key -e OPEN_AI_MODEL=gpt-4-0125-preview -e OLLAMA_MODEL=llama2 -p 3000:8080 -v /var/run/docker.sock:/var/run/docker.sock ghcr.io/semanser/codel:latest
• Prerequisites: OpenAI API key or a locally hosted Ollama model.
• Resource Footprint: Requires Docker and sufficient resources for running LLMs and the agent.
• Links: Discord, Roadmap

Highlighted Details

• Secure, sandboxed Docker environment.
• Built-in browser for web information retrieval.
• Integrated text editor for file viewing and modification.
• Automatic Docker image selection based on task.
• PostgreSQL for command history and output persistence.

Maintenance & Community

The project has a Discord server for community interaction. Key inspirations include research papers on autonomous agents and projects like Devin.

Licensing & Compatibility

The repository's license is not explicitly stated in the README. Compatibility for commercial use or closed-source linking is not specified.

Limitations & Caveats

Codel requires specific environment variables for LLM integration (OpenAI or Ollama). The README does not detail specific hardware requirements beyond Docker, nor does it mention any known limitations or alpha/beta status.