BeeBot is an autonomous AI agent designed to perform practical tasks, targeting users seeking automated assistance. The project aims to provide a flexible and reliable system for task execution, though its development is currently on hold due to limitations in current LLM capabilities for generalized autonomous AI.

How It Works

BeeBot utilizes a tool-selection mechanism via AutoPack, allowing it to acquire and use tools dynamically during task execution. It employs LLMs, primarily OpenAI, for planning and decision-making, with a philosophy of tool-focus and LLM specialization to ensure reliable task completion. The system prioritizes functionality and flexibility, employing unorthodox methodologies like the absence of unit tests in favor of end-to-end testing.

Quick Start & Requirements

• Install dependencies using poetry after cloning the repository.
• Persistence is required; PostgreSQL via Docker is recommended (docker compose up -d).
• CLI usage: poetry run beebot
• API usage: Start server with uvicorn beebot.initiator.api:create_app --factory --timeout-keep-alive=300.
• Official documentation is available in the docs/ directory.

Highlighted Details

• Built-in persistence with swappable filesystem emulation (in-memory, disk, database).
• REST API conforming to the e2b standard.
• Websocket server for event notifications.
• Caching with Helicone (if enabled).

Maintenance & Community

Development is currently on hold, with a plan to revive the project when LLMs improve or a custom model is developed. The project welcomes contributions and feedback.

Licensing & Compatibility

The repository does not explicitly state a license in the provided README. Compatibility for commercial use or closed-source linking is not specified.

Limitations & Caveats

Windows is officially unsupported. Notifications via the websocket server are undergoing rework and may not be functional. The project is in development, with a stated absence of unit tests and a focus on end-to-end testing.