Havuz

Give AI agents fast, bounded access to a real Android device or emulator. MCP server and CLI for Claude Code, Codex, Cursor, and more.

Automate your real iPhone from Safari with a $2 ESP32. No Mac, no APP, and no jailbreak. Local conversations, your choice of AI, and communi

Python desktop AI assistant for screen awareness, voice interaction, persistent memory, and computer automation

Computer use for the Pi coding agent — Windows apps, browsers and desktops through native UI trees, screenshots and real input.

让 Agent 操控 Windows 桌面软件

Voice-first Mac assistant that uses the computer for you. GPT-Live-1 voice behind a wake word; brains: Codex, Claude Code, an API or a local

基于pi-agent自己开发的一个agent desktop工具

Multi-modal benchmark for measuring sensitive-information redaction in human recording data

Verified control of the whole desktop for agents on Windows and Linux: apps with no API, windows, shell, browser. MCP server + CLI + REST AP

The desktop AI agent for the real web. Built-in fingerprint browsers, saved logins, multi-agent orchestration, and reusable Browser Skills b

Fast browser and macOS computer use for MCP agents. Local execution, TypeSafe Jev decisions, verified outcomes.

Analyze jailbroken iOS and macOS apps, then build and verify independent implementations of selected behavior.

Open-source local-first computer-use agent for Linux, Windows, and macOS, powered by local VLMs or compatible APIs.

Open-source Computer Use MCP server for any coding agent — macOS, Windows, Linux. Accessibility-first, background input, agent pointer, zoom

Self-hosted AI agents with persistent memory, browser and computer access, and scheduled workflows.

Local, open-source Jev alternative: browser agent decisions with Laya (System One model) on your own machine. No cloud, no API key. Playwrig

AI screen companion for Windows 11 — voice-controlled, sees your screen, points to what you ask for. Windows port of github.com/farzaa/click

Windows computer-use MCP server: drive any desktop app via semantic discover-then-act targeting (entities + leases, not pixel coordinates),

Open-source macOS AI computer use and native GUI automation on Apple silicon. Jev + OmniParser CoreML + Apple Vision OCR. Bring your own Ope

Six-skill spec-driven engineering pod for Claude Code — scaffold specs, research, discover direction, generate design variants, implement, a

Remote Desktop Control for AI agents: screenshot, mouse, keyboard and window control of another machine over Tailscale, exposed as MCP tools

Self-improving agent harness for ChatGPT and Codex on Mac. #1 on AssistantBench. On-disk memory, privacy zones, verified completion.

Let agents work in your apps while you keep using your Mac.

Cross-platform native desktop control for DeepSeek Harness (Windows first): screen_shot, screen_read, click/type/scroll/key, app_list/app_la

Drive your Mac focus-free over MCP: OS APIs, AppleScript, CDP, and Accessibility

Drive a GNOME Wayland desktop from AI agents. CLI-first, no portals, no consent dialogs.

Cross-platform Rust computer-use MCP server for macOS and Windows: AX/UIA, screenshots, OCR, pointer and keyboard input, clipboard, stdio, a

macOS computer-use agent in the notch. Long-press, talk, Claude drives the mouse.

Real-world browser-agent benchmark: 210 tasks across 107 websites, multi-agent/multi-browser evaluation, reproducible leaderboard and result

Native macOS coding agent and AI coworker with local tools, Computer Use, worktrees, automations, plugins, and verified auto-updates.

On-device iPhone automation for AI agents. Open-source Swift SDK for screenshots, taps, swipes, text input and an embedded local tunnel.

Train computer-use agents end-to-end on real desktops, at scale: a high-performance runtime where 8K+ live environments on one laptop thread

A streaming action language for computer-use agents. LLMs emit one JSON action per line; the executor runs each the moment its newline arriv

把 AI 接入你的真实开发环境。CodePier 让 ChatGPT / MCP 安全连接本地项目、Pi / Codex、Shell、浏览器与桌面能力。

Reverse-engineering & re-implementation of OpenAI Computer Use macOS plugin for DeepSeek Harness: 10 MCP tools, JSON-RPC wire protocol, nati

Ekbasis: an open world model for agents. What an action will do, before it is done.

AI vision GUI automation for browsers, mobile apps, desktops, and games — no DOM, no selectors. Standalone or on top of Playwright / Seleniu

Drive Blender from AI agents with a plain Python SDK (no MCP), grounded in REA/Ghidra reverse engineering of Blender 5.2.2

A browser for LLMs, not humans.

Open-source desktop coding agent for macOS. Use any model: local with Ollama, your own API keys or a ChatGPT sign-in. Parallel chats, MCP co

DSH 插件：在权限菜单里新增「管理员」档（完全访问 + 永不询问），内置官方 computer use、操控提示与自主迭代按钮。

Hybrid computer use for AI agents on macOS.

Open-source alternative to Grok bot with a FastAPI API, Next.js UI, governed computer use, Docker/remote runtimes, approvals, connectors, an

An assistant for Windows that works inside 118 desktop programs and services through code instead of screenshots. It writes each program's o

Local-first, IM-native AI coworker runtime with async role agents, trace replay, evidence, and agentic evaluation.

Shellby MCP turns ChatGPT Web into a local coding agent with full computer access, persistent tools, and multi-agent capabilities.

CCP（Claude-Codex-Pro-Tool）｜一个窗口管好你所有的 AI 编程 Agent：Token 用量看板、第三方 API 一处切换、会话蒸馏成 Skill、Claude Desktop 的 Computer Use。Rust + Tauri，数据只在本机。 Eng

Synthetic users that walk your software before real users do.

Download, extract and open Hélène. A desktop home for independent agents with lasting memory, tools, Telegram and optional local voice. Powe

Local computer-use host: agents drive your own macOS or Windows desktop over MCP stdio, with your own model keys and kill switch, allowlist