Report Abuse

Basic Information

BeeBot is an open source autonomous AI assistant platform intended to run and manage practical tasks autonomously and to provide developers with the infrastructure to build and operate task-oriented agents. The project supplies a task model, an API surface, persistence, event streaming, and tool management so an LLM-driven agent can select and use external tools during task execution. The README emphasizes BeeBot as both an autonomous worker and a developer-facing system with a CLI, a REST API conforming to the e2b standard, and websocket notifications. Persistence is required and Postgres is recommended. The project is currently on hold as of late 2023 pending improvements in LLM capability or a specialized model, and documentation is light but present in a docs/ directory. The repo focuses on practical execution, tool selection, and integration rather than on complex memory or tree-of-thought architectures.

Links

App Details

Features
BeeBot provides tool selection via AutoPack and can acquire additional tools during task execution. It includes built-in persistence with official SQLite support for tests and recommends running Postgres for production. The system exposes a REST API matching the e2b standard and a websocket server that publishes events and model changes. BeeBot supports a swappable filesystem emulation layer to store files in-memory, on-disk, or in a database. There is built-in caching integration with Helicone if enabled. The project ships a CLI entry point and documents uvicorn-based commands to run the API server. Other features include dynamic manipulation of task history, plans for a web UI built with Remix/Node.js, and emphasis on end-to-end testing.
Use Cases
For developers and teams building autonomous agents BeeBot supplies a ready-made orchestration layer that handles task creation, stepwise execution, tool orchestration, persistence, and event streaming. Its e2b-compatible REST API and websocket notifications make integration with external systems straightforward. The swappable filesystem and persistence options simplify handling artifacts and long-running workflows. Built-in caching and tool selection reduce repetitive work and make LLM-driven tooling more practical. The CLI and documented API examples demonstrate how to create tasks and execute steps, enabling rapid prototyping and testing of agent behaviors. The README cautions that development is paused pending better LLM reliability, and it recommends Postgres for durable usage; documentation is light but extensible via the docs/ folder.

Please fill the required fields*