LobsterAI is an open-source desktop agent that runs on local files and workflows, integrates with multiple IM platforms, and uses OpenClaw for agent execution. The project is TypeScript-based, MIT licensed, with a sizable community and ongoing updates.
Collecting history — the radar snapshots this repo daily. The trend line appears after 3 days of data (1 so far).
What it is
LobsterAI is an open-source, desktop-grade AI agent that can operate in local environments (files, terminal commands, browser workflows, documents, slides, and more) and can be controlled via IM channels. It uses a split architecture: Desktop Cowork sessions and OpenClaw runtime underneath. It supports multi-agent workflows, expert kits, and a set of built-in skills. It provides local memory via SQLite and artifacts rendering inside the app. It binds to various IM gateways (WeChat, WeCom, DingTalk, Feishu/Lark, Telegram, and more).
How it works
The architecture includes a Renderer (React, Redux Toolkit, Tailwind) for UI, a Main process (Electron) for lifecycle, IPC, and persistence, and an OpenClaw integration layer that translates LobsterAI state into runtime behavior. It exposes MCP servers for connecting external tools and data sources, and uses local worker memory files like MEMORY.md, USER.md, and SOUL.md for context.
Getting started
Install guidance covers desktop download from Official Website or GitHub Releases, and running from source:
git clone https://github.com/netease-youdao/LobsterAI.git
cd LobsterAI
npm install
First development run:
npm run electron:dev:openclaw
Daily development after the pinned OpenClaw runtime exists:
npm run electron:dev
Renderer dev server runs at http://localhost:5175.
OpenClaw Runtime
# Build the current-platform runtime manually
npm run openclaw:runtime:host
# Use a custom OpenClaw source checkout
OPENCLAW_SRC=/path/to/openclaw npm run electron:dev:openclaw
# Force runtime rebuild
OPENCLAW_FORCE_BUILD=1 npm run electron:dev:openclaw
# Keep a local OpenClaw checkout on its current branch/tag
OPENCLAW_SKIP_ENSURE=1 npm run electron:dev:openclaw
Recent releases
Latest updates include features such as:
- cowork text tagging in side chat and support for KIMI K3 integration
- improved AI skin creation flow
- surface of structured run failure details in error UI
- various refactors and enhancements in cowork run handling and UI
Traction
Star count is shown as 5769; forks: 905; open issues: 758.
Packaging and distribution
Packaging commands to build installers for macOS, Windows, and Linux are provided, including options for portable Python and environment-specific packaging
npm run dist:mac
npm run dist:mac:x64
npm run dist:mac:arm64
npm run dist:mac:universal
npm run dist:win
npm run dist:linux
License
MIT