beebot

An Autonomous AI Agent that works

open-sourceagent-frameworks
Visit WebsiteView on GitHub
452
Stars
+38
Stars/month
0
Releases (6m)

Overview

BeeBot is an autonomous AI assistant designed to perform practical tasks without human intervention. Built as a 'personal worker bee,' it can dynamically select and acquire tools during task execution through AutoPack integration. The system features a modular architecture with swappable filesystem emulation, built-in persistence using SQLite or PostgreSQL, and a REST API conforming to the e2b standard. BeeBot includes event publishing via websockets, dynamic history manipulation, and optional caching through Helicone. However, development is currently on hold as of late 2023, with the creator citing current LLM limitations in structured thinking and reliable autonomous operation. The project may resume if LLMs improve significantly or if a specific subset of tasks proves viable for autonomous execution.

Pros

  • + Modular architecture with swappable filesystem emulation and multiple storage options
  • + Comprehensive API ecosystem including REST endpoints, websockets, and e2b standard compliance
  • + Dynamic tool acquisition and selection capabilities through AutoPack integration

Cons

  • - Development currently on hold due to perceived LLM limitations for autonomous tasks
  • - Windows officially unsupported with potential compatibility issues
  • - Requires mandatory persistence setup and PostgreSQL recommended for production use

Use Cases

Getting Started

1. Clone the repository and run './setup.sh' to install dependencies via poetry 2. Start PostgreSQL database using 'docker compose up -d' for persistence 3. Launch BeeBot via CLI with 'poetry run beebot' or start the API server with uvicorn