MCP servers for computer-use

Desktop automation MCP server — computer use for any AI agent: control screen, windows, mouse/keyboard, and Chrome via Model Context Protocol (stdio)

System 1 decision models (Jev, Laya, Cua-S1) as brain for agents: Browser use, computer use, games and robotics

A floating chat bubble that lets Claude Code, Codex or OpenCode use your Android phone. Runs inside the app on your own subscription or free models. No adb, no PC, no root.

GenPark AI Agent Skill - Grounds natural language browser instructions to precise (x, y) coordinate click bounding boxes and interactable targets.

GenPark AI Agent Skill - Intercepts and auto-dismisses intrusive GDPR/CCPA cookie consent banners, paywall overlays and shadow-DOM modals.

GenPark AI Agent Skill - Compresses raw HTML DOM trees into token-efficient semantic accessibility representations with interactive element IDs.

GenPark AI Agent Skill - Validates and normalizes browser WebGL, Canvas, and Navigator fingerprint attributes to prevent automated detection.

GenPark AI Agent Skill - Resolves multi-step cascading form fields and conditional dynamic dropdowns with event propagation.

GenPark AI Agent Skill - Validates and normalizes browser WebGL, Canvas, and Navigator fingerprint attributes to prevent automated detection.

GenPark AI Agent Skill - Intercepts and auto-dismisses intrusive GDPR/CCPA cookie consent banners, paywall overlays and shadow-DOM modals.

GenPark AI Agent Skill - Grounds natural language browser instructions to precise (x, y) coordinate click bounding boxes and interactable targets.

GenPark AI Agent Skill - Resolves multi-step cascading form fields and conditional dynamic dropdowns with event propagation.

GenPark AI Agent Skill - Compresses raw HTML DOM trees into token-efficient semantic accessibility representations with interactive element IDs.

Vision-to-DOM element anchor resilient against click drift (Genspark Claw / Computer Use)

Multi-agent debate cross-examination & consensus arbitration (Super Agent / DeepSeek-R1)

Sub-100ms conversational turn-taking & barge-in handler (Cartesia Sonic / ElevenLabs)

Sub-100ms conversational turn-taking & barge-in handler (Cartesia Sonic / ElevenLabs)

Autonomous merchant checkout token & velocity limit guardrail (Agentcard / Stripe)

Heterogeneous search synthesis to interactive comparison layout (Genspark Sparkpages)

Multi-agent debate cross-examination & consensus arbitration (Super Agent / DeepSeek-R1)

Vision-to-DOM element anchor resilient against click drift (Genspark Claw / Computer Use)

Heterogeneous search synthesis to interactive comparison layout (Genspark Sparkpages)

Autonomous merchant checkout token & velocity limit guardrail (Agentcard / Stripe)

Geometric spatial relationship grounding evaluator verifying relative positional assertions between objects and UI controls

Set-of-Marks (SoM) bounding box resolution and relative viewport coordinate mapper for computer-use vision agents

Set-of-Marks (SoM) bounding box resolution and relative viewport coordinate mapper for computer-use vision agents

Multi-frame GUI computer-use action trajectory summarizer detecting dead clicks and UI rendering latency stalls

Browser DOM accessibility tree pruner filtering non-interactive noise and generating dense actionable trees for web agents

Multimodal visual hallucination verifier cross-checking vision model claims against ground-truth OCR and bounding box metadata

Multi-frame GUI computer-use action trajectory summarizer detecting dead clicks and UI rendering latency stalls

Browser DOM accessibility tree pruner filtering non-interactive noise and generating dense actionable trees for web agents

Geometric spatial relationship grounding evaluator verifying relative positional assertions between objects and UI controls

Multimodal visual hallucination verifier cross-checking vision model claims against ground-truth OCR and bounding box metadata

Computer use for Hyprland - MCP server giving AI agents native hands on your Wayland desktop. No ydotool, no root, no portals.

Vision-primitives MCP server for text-only LLMs: describe, locate (coordinates), OCR with bbox, annotate, crop, zoom, automated anomaly scanning, computer use — powered by Xiaomi MiMo V2.5. Single-fil

Computer use for your agents across the Omarchy machines you own

Windows computer-use MCP server (26 tools) with a human-in-the-loop approval gate: physical-input-only dialogs, signed guard switches, deny/launch lists, hash-chained audit log. 70+ unit tests, green

Browser action recorder for automation: record a task once, replay it, or hand the folder to an agent

The Open-Source Astra 6 for your PC. Your AI can see, click, type, navigate, and get real work done on Windows.

AX-first computer use for AI agents on macOS with optional Jev (TypeSafe System One) semantic guards: calibrated target/input judgments before an irreversible action, decisions kept in code. Accessibi

MCP server for AI agents to observe and drive a niri Wayland desktop. Early development.

DeepSeek Harness (DSH) plugin: mount the nuphus-mcp desktop automation MCP server as native tools (mcp__nuphus-mcp__*).

🖱️ Let Factory Droid control your Mac with the Codex app's Computer Use tool. One-command install, no approval forms, works with any Codex install path.

Windows desktop automation MCP server - screenshots, OCR, mouse and keyboard, and accessibility trees via UI Automation, MSAA and Chrome DevTools Protocol. Node.js + built-in PowerShell, no Python and

Leap gives any AI agent x-ray vision into desktop apps (native accessibility and screenshots) along with full mouse and keyboard control. It works in the background without taking over your screen, us

Give your AI its own Windows desktop session while you keep using yours. Local viewer, CLI and MCP. Experimental preview for Windows 11.