LobsterAI - Desktop AI Agent

LobsterAI is a desktop-grade AI Agent launched by NetEase Youdao in 2026, and the first open-source desktop-level Agent released by a major Chinese tech company. It runs directly in your real desktop environment, capable of operating local files, executing terminal commands, controlling browsers, processing documents and spreadsheets, generating presentations, creating videos, and more — truly "one-click install, ready to use."

Core Features

• Desktop Cowork Sessions: Supports long-running Agent tasks that deeply interact with local projects and files. LobsterAI streams progress in real time, preserves session history, renders tool output, and requests user approval before sensitive actions such as file operations, terminal commands, or network access.
• Multi-Agent Workflows: Allows users to create multiple custom Agents, each with its own identity, model selection, skill configuration, working directory, and IM bindings. The main Agent handles general tasks while specialized Agents serve repeatable, dedicated roles.
• Expert Kits: Provides scenario-oriented Expert Kits that package capability selections and references into ready-to-use solutions for common workflows. Kits can be freely combined with individual skills for flexible task handling.
• Rich Skills Library: Ships with 28 built-in skills covering web search, Word documents, spreadsheets, PowerPoint, PDF processing, Remotion video generation, browser automation, image/video generation, stock research, content writing, email, weather, and skill creation.
• MCP Server Integration: Connects external tools and data sources through the Model Context Protocol, with user-configured servers stored locally and synced to the OpenClaw runtime.
• Scheduled Tasks: Supports creating recurring work through conversation or the scheduled task UI, ideal for daily news digests, inbox summaries, website monitoring, weekly reports, and other repeatable tasks.
• IM Remote Control: Reach your desktop Agent from WeChat, WeCom, DingTalk, Feishu/Lark, QQ, Telegram, Discord, NetEase IM, POPO, and email. Multi-instance platforms can bind different accounts or channels to different Agents.
• Rich Artifacts Preview: Preview and manage generated HTML, SVG, images, video, Mermaid diagrams, code, Markdown, text, documents, and local service artifacts within the desktop app.
• Local Memory & Data: Sessions and app data are stored locally in SQLite. OpenClaw workspace memory uses files such as MEMORY.md, USER.md, and SOUL.md for persistent preferences and cross-session project context.

Technical Architecture

LobsterAI is built on the Electron desktop framework, with a React + Redux Toolkit + Tailwind CSS rendering layer. The main process handles Electron lifecycle, IPC communication, SQLite persistence, authentication, logging, and OpenClaw runtime management. Its underlying runtime, OpenClaw, is an independent Agent gateway. LobsterAI translates product state into runtime behavior through the OpenClaw engine manager, config syncer, and runtime adapter. The project has garnered over 5,400 GitHub Stars, is licensed under MIT, and supports both macOS and Windows platforms.

Typical Use Cases

• Build a local inventory management system
• Analyze local data and generate visual dashboards
• Research AI market trends and create presentations
• Automatically check ad spend and detect anomalies daily
• Batch-screen resumes and generate candidate shortlists
• Collect and summarize news on weekday schedules

GitHub: https://github.com/netease-youdao/LobsterAI