eigent-ai/eigent

Eigent is an open source desktop application for building and running a team of AI agents that work together to complete tasks. The README calls this team a workforce: rather than one assistant doing everything, Eigent splits a job into parts and assigns them to several specialized agents that run at the same time. It describes itself as a free and local alternative to a commercial product called Claude Cowork, and it is built on top of an existing open source project named CAMEL-AI. The README lists the agents that come ready to use. A Developer agent writes and runs code and terminal commands, a Browser agent searches the web and pulls out information, a Document agent creates and manages documents, and a Multi-Modal agent works with images and audio. The agents can be given extra tools through something called the Model Context Protocol, or MCP, which connects them to services such as web browsing, Notion, the Google suite, and Slack, and you can add your own tools as well. If an agent gets stuck or is unsure, Eigent can pause and ask a person for input. There are a few ways to run it. The recommended option is local deployment, where the whole system runs on your own machine with your choice of local AI models and no cloud account, keeping your data under your control. There is also a quick cloud-connected mode that uses Eigent's servers and requires registration, plus a managed cloud version and an enterprise offering with features like single sign-on and access control. For the cloud-connected setup the README gives the commands: you need Node.js, then you clone the repository, install the dependencies, and start it in development mode. The project is written with a TypeScript frontend and a Python backend. The whole codebase is open source, so anyone can download, read, and change it, and the README points to guides, community channels, and a roadmap for following the project's progress.