LobsterAI
Open-source, desktop-grade AI agent that gets real work done — data analysis, slides, docs, video & web research. Built on OpenClaw; runs tools on your real des
What is LobsterAI?
Open-source, desktop-grade AI agent that gets real work done — data analysis, slides, docs, video & web research. Built on OpenClaw; runs tools on your real desktop and takes commands from your phone via WeChat, Feishu, DingTalk & Telegram.
Overview of LobsterAI
What it is
LobsterAI is an open-source, desktop-grade AI agent developed by NetEase Youdao. It is designed to operate directly on a user's real desktop environment, executing tasks such as data analysis, document creation, slide generation, video production, and web research. The agent runs tools locally and can be controlled remotely via instant messaging platforms including WeChat, Feishu, DingTalk, and Telegram. LobsterAI uses OpenClaw as its runtime and gateway, while its own product/session layer is called Cowork. The separation allows LobsterAI to maintain local persistence, permissions, UI state, artifacts, agents, memory, and IM bindings in the desktop app, while delegating agent execution to OpenClaw.
Key features
- Desktop Cowork Sessions: Run long-form agent tasks against local projects and files, with progress streaming, session history, tool output rendering, and approval prompts for sensitive actions.
- Multi-Agent Workflows: Create custom agents with their own identity, model choice, skills, working directory, and IM bindings; use specialized agents for repeatable roles alongside a main agent.
- Expert Kits: Install scenario-oriented kits that package capability selections and references for common workflows, combinable with individual skills.
- 28 Built-in Skills: Includes web search, document creation, spreadsheets, PowerPoint, PDF processing, Remotion video generation, browser automation, image/video generation, stock research, content writing, email, weather, and skill creation.
- MCP Servers: Connect external tools and data sources through Model Context Protocol servers; user-configured servers are stored locally and synced into OpenClaw.
- Scheduled Tasks: Create recurring work via conversation or scheduled task UI for daily news digests, inbox summaries, website monitoring, weekly reports, and more.
- IM Remote Control: Control the desktop agent from WeChat, WeCom, DingTalk, Feishu/Lark, QQ, Telegram, Discord, NetEase IM, NetEase Bee, POPO, and email.
Use cases
The README provides real-world example prompts including: building a local inventory system from Excel data, analyzing local data files to create visual dashboards, researching a market topic and generating a presentation, automating daily browser checks on ads dashboards, screening resumes against a job description, and scheduling daily AI news digests at 9 AM.
Getting started
Download the latest macOS and Windows installers from the official website or GitHub Releases. To run from source, Node.js >=24.15.0 and pnpm >=11.17.0 are required. See the repository README for detailed instructions.
License and pricing
License: MIT. The project is open-source.
Based on the project's public README and repository metadata.