Computer use & desktop agents
727 agents · most starred first
Computer use & desktop agents, page 5 of 8
Open-source AI operating system for computer use agents
AI desktop agent with full PC control. Single Go binary + Tauri desktop app.
Automate QQ Farm tasks with AI vision: harvest, weed, water, plant, and steal crops with CV-based bot control
Open-source Android computer-use agent with a self-hostable backend
Menu bar app that types your script into any app - one block per hotkey press, as real keystrokes at a human pace. For demos, screencasts, and presentations.
Lightweight Wayland scaffold for AI agents to operate one GUI app without stealing the human's host focus, cursor, workspace, or clipboard.
Give your AI an eye — local screen → vision-model captions → memory store your AI reads on wake-up.
AI that controls your Mac. Describe a task, watch it happen. Native macOS desktop app powered by Anthropic Computer Use.
Give any AI agent general-purpose computer use: live screen perception (OCR, vision frames, MJPEG stream) plus safety-gated mouse, keyboard, window, virtual-desktop and game-mode control over a local HTTP API and MCP.
Select the local project folder on the AI website so that the AI can directly read and modify files. Supports ChatGPT/Claude/Gemini/DeepSeek/Kimi/DouBao/YuanBao.在 AI 网站上选择本地项目文件夹,让 AI 直接读取和修改文件。支持 ChatGPT/Claude/Gemini/DeepSeek/Kimi/豆包/元宝。
GenPark skill: Utilize GenPark Workspace 4.0 for desktop automation, MS Office integration, and AI meeting notes.
AI assistant with a local Linux desktop, live browser view, and task replay.
A grounded, local-first computer agent powered by Google Gemini with safe tools, bounded autonomy, and verifiable actions.
Background computer use for Windows: coding agents (Codex, Claude Code, any MCP client) drive GUI apps on a hidden desktop with real mouse and keyboard while you keep using your PC. Each agent gets its own layer.
Hand desktop creativity over to your AI agent
Open-source Computer Use in any agent.
AI-native Linux desktop where your AI agents (Claude Code, Hermes, OpenClaw, Pi) drive every app, remember with YantrikDB, and show proof of what they did. Boot the live ISO.
Windows desktop automation for AI agents. MCP server and CLI with a background child session, screenshots, separate mouse/keyboard focus and app testing.
Open Source, Self Hosted, AI Visual workflow builder for automating anything Desktop running inside its own Virtual Computer.
An AI-native desktop environment
A computer-use agent that grows its own team: it sees your screen, clicks and types on your real desktop, and clones specialists and forges skills on demand.
A desktop AI agent that controls your local machine — runs commands, manages files, executes code, browses the web autonomously etc. Supports Claude, GPT, Gemini, Llama, DeepSeek, and more. .exe available
LLM-powered macOS computer control. AgentPilot lets AI agents see your screen, move the mouse, type text, press hotkeys, manage windows, run shell commands, and more — all through a native macOS app with a chat interface.
面向中文开发者和中文用户的通用 AI agent 访问层:把网站、App、API 和重复流程路由到 agent-native CLI。
Gesture + voice control for macOS. Pinch to move cursor. Swipe for arrows. Circle to screenshot. Speak to run commands.
🚀 PE-SmartFixer:国内首款基于 AI 视觉大模型(Qwen-VL)的 Windows PE 离线系统修复与蓝屏诊断智能工具。纯 Go 语言编写,支持脱机维护与注册表挂载。
GenPark skill: Utilize GenSpark Workspace 4.0 for desktop automation, MS Office integration, and AI meeting notes.
ZYRAXON AI — The World's Most Advanced Desktop AI Agent.
Screen-aware desktop agent that turns voice, screenshots, docs, and current context into live Gemma 4 + Cerebras task graphs.
A private macOS virtual machine on your Mac, operated by an AI agent. You chat; it clicks, types and hands back checked results. Also a CLI and MCP server for coding agents.
Vision-first desktop computer use for DSH on macOS: OCR text targeting, set-of-mark indices, 11 desktop_* tools and a visible cursor. No accessibility tree. · 让 DSH 看着屏幕操作你的 Mac。
RaspberryPi KVM and Agent AI Control for Modern and Retro Computers
Free, open-source AI assistant for macOS. Talk about your screen, get spoken step-by-step guidance and visual answers, and delegate tasks to specialist mates. Native SwiftUI; bring your own Claude, ChatGPT or compatible API.
Windows computer-use MCP server for dsh (DeepSeek Harness) — client-neutral, works with any stdio MCP client. Accessibility-first element actions, skyshot text trees, guarded raw input, and a cross-session arbiter that yields to the human.
Computer Use for Agentic Process Automation
Anthropic's Computer use implementation in Nodejs
MyCodex desktop AI agent workspace for Codex + Hermes workflows, WeChat control, browser automation, file artifacts, and token-light setup.
The AI companion that actually sees your screen. Debug code visually, watch videos together, get help without copy-paste.
Drive Electron apps the way Playwright drives browsers — MCP server for AI agents.
Build faster with an open-source AI workspace designed for real developer workflows.
A high-performance desktop app interaction framework for AI agents — faster, cheaper, more accurate. / 让 Agent 操作桌面软件更快、更准、更省。
Governed AI agents that actually do the work - in the cloud or on your own machine - across business systems, real browser tabs and real Android apps. Visual builder, MCP server, and official SDKs in 9 languages.
Computer Use Examples with API Integration Full Stack
gradio region:us
面向 AI Agent 的桌面 UI 协议与控件库,支持 WPF / WinForms 控件识别、高亮与本机语义远程控制。
The Open-Source Computer Use Engine for your PC. Your AI can see, click, type, navigate, and get real work done on Windows.
Windows computer-use MCP server - 1:1 replica of Claude Desktop's tool surface, for the Claude Code CLI or any MCP client. Screenshots, mouse, keyboard, multi-monitor. Full tier everywhere: type into IDEs, launch loose dev builds, let the agent test the app you're building.
Windows UI Automation extension of arc-cua: a fast, JEV-powered decision loop for desktop computer-use agents
面向 AI Agent 的桌面 UI 协议与控件库,支持 WPF / WinForms 控件识别、高亮与本机语义远程控制。
Your computer work. Done for you. The AI coworker for everyone who works at a computer: tell it the task in plain words, it signs in, clicks through and finishes the job in your own browser and apps. Free for 14 days on Mac. MIT.
HeyFlitty: a free, open-source AI screen buddy for Windows (and macOS). Hold a hotkey, ask out loud, and it answers, captions it, and flies a cursor to the exact button you need.
Cross-platform GUI MCP for interacting with your desktop in macOS, Windows, and Linux!
12 个主流 Coding Agent 的 Computer Use / Browser Use 能力逆向图谱:ZCode/Codex/Claude/Cursor/MiniMax/Synara/Kimi/Qoder/Grok/Devin/Goose/MiMo——架构、工具面、安全模型与可复用源码。在线主站:https://freeall12.github.io/computer-use/
WinReach — give your AI agent its own Windows machine. Remote PowerShell, file transfer, screenshots, and full desktop control (computer use) over HTTP, secured with per-user keys, roles, command policies, mTLS, and an audit log. A Model Context Protocol (MCP) server for Windows.
Free, open source hand-gesture and voice control for Windows. Webcam gestures move the cursor; say Zehen and an AI assistant opens apps, clicks, types and answers out loud. Offline wake word, no extra hardware.
An open-source, instantaneous, and superior alternative to GenPark Workspace 4.0 Desktop AI integration.
Background screen and meeting context observer.
Geometric spatial relationship grounding evaluator verifying relative positional assertions between objects and UI controls
A second cursor for Claude Code, Codex, Cursor and other agents: it points, explains and walks you through your Mac. Hold ⌃⌥ to circle anything and ask. Open source, MIT.
Pre-built automation workflows for OpenOwl + Claude Code. Clone, configure, run.
ACI (Agent-Computer Interface) — The open standard that turns any application into a structured, agent-operable interface. One protocol. Any software. No screenshots.
Playwright but for any app on any OS
Interactive demo — AI agent filling forms via ACP (Agent Control Protocol)
Computer Use, App Use, AppBot — give AI agents real eyes and hands on Windows. Open-source RPA where humans, the app ecosystem, and AI share one keyboard. Focusless. Self-healing. Multi-AI.
Autonomous AI desktop agent powered by local Ollama LLM — automates files, Excel, web, shell and GUI tasks on Windows.
Portable Desktop AI Agent. Built on LangGraph & MCP. Safe by design: human-in-the-loop for all critical actions. Self-correcting, api-key-rotation, multimodal, multi-tool coding assistant in a single .exe.
A desktop environment where agents do the work.
let poke use your computer, with macos, linux and windows support.
一个会读写代码、操控终端与网页的桌面 AI 智能体(Electron + Python)。A desktop AI agent that reads your code, runs your terminal, and drives the web.
Open a browser tab — you're now driving your computer. Live remote desktop + real PTY shell (Claude Code, vim, top) + on-device voice. macOS and Windows, every machine you own behind one URL and one password. Self-hosted.
An in-QGIS agentic chat assistant. Type a request in plain language and an LLM agent carries it out inside the running QGIS session If QGIS can do it, the agent can do it: anything you could script in the Python console or run from a toolbox, it reaches the same way.
Teach once, replay anywhere — harness-agnostic Teach-by-Demonstration engine (Record → Compile → Replay → Verify) with OpenMausBot and DeepSeek Harness integrations.
Auditable vision and cross-platform Computer Use runtime for DeepSeek Harness — strict evidence, health-checked failover, original pixels, and Token accounting.
Open-source AI agent that does real work on your Windows PC - browser, files, apps and accounts - while you keep using it. Send it tasks from your desk, phone, WhatsApp or Telegram.
General-purpose autonomous task decomposition, multi-modal web/sandbox execution loop, and self-correcting goal resolution (inspired by Manus AI).
Ambient desktop context listener, screen text & clipboard change trigger, and non-intrusive proactive intent predictor (inspired by Cue).
AIOPE — 70-tool on-device AI agent for Android. Linux terminal, realtime voice, browser automation, SSH, on-device RAG, dynamic UI, MCP. BYOK, any model. By XNet Inc.
🖱️ AI Mouse: ask your Windows screen · 🛡️ Hardened to BSI requirements (IT-Grundschutz APP.3.1, TR-02102) · Self-hosted autonomous AI agent for Linux — multi-LLM, real desktop control (VNC), WhatsApp, RAG knowledge base, sandboxed security layer, OpenClaw skills
Visual desktop automation agent — low-code workflow builder with AI-powered steps
Autonomous AI Agent for Windows - Your intelligent, always-on AI assistant with screen understanding and task automation.
Your screen history, searchable and automated - Windows OCR + REST API + AI + RAG
Official implementation for "UI-Hawk: Unleashing the Screen Stream Understanding for GUI Agents" (EMNLP 2025)
ComputerMate is a powerful Model Context Protocol (MCP) server that grants AI models the ability to interact with your computer just like a human would.
Open-source Marvis-like desktop AI agent for macOS — Main + File + Search Agents, layered dispatch protocol, mv-* card rendering, 3-tier security gate. FastAPI + Next.js. Apache 2.0.
Mark XLVIII is a cross-platform, real-time voice AI assistant built on the Gemini Live API that can hear, see, and understand you. It features visual awareness, autonomous task planning, and direct OS-level control to serve as an intelligent, hands-free extension of your digital workspace, So follow up to know and see new models.
A native macOS AI assistant that sees your screen, remembers context per app/site, and can guide you click-by-click through on-screen tasks and smartly answer any queries from the web, if any.
Zero-dep macOS desktop automation MCP server for AI agents — 18 tools, macOS 13+. npx mac-use-mcp
OpenClaw, but safer and better. Optimized for whatever app you have.
A lightweight, hackable AI agent that can control your machine by executing Bash scripts
Playback Electron App
TypeScript SDK for the Computer Use Protocol
A powerful desktop assistant built in Python that combines voice commands, GUI interface, and system automation capabilities.
APIfy your local workspaces, allowing AI to interact with it.
无糖管家-语音控制的电脑管家,几乎可以做任何事。
Computer Experience Layer — MCP server for native computer use on macOS. Structured perception, reliable actions, continuous awareness.
Use an LLM agent to automate ordering food and other items from Deliveroo, Uber Eats, DoorDash, etc.
IEE is a native C++ system runtime that converts live OS/application state into a deterministic, queryable, and executable intent layer. It is built as an execution substrate and control plane, not a UI product.
🤖 Control your Gemini computer effortlessly with a unified console for browser automation, desktop management, and intelligent task execution.
Screenshot CLI that frames any window into a polished, share-ready card and prints stable JSON for shells, hotkeys, and AI agents