Skip to main content
hc
Submit
mcpSource-backed

Windows MCP Server

Windows computer-use MCP server that lets AI agents interact with the Windows operating system through UI state, keyboard and mouse actions, application control, file navigation, QA testing, and browser DOM mode.

by CursorTouch · submitted by oktofeesh1·added 2026-06-05·
Claude CodeCodexCursorClaude Desktop
HarnessClaude CodeCodexCursorClaude Desktop
Review first — review before installing

Open the source and read safety notes before installing.

Citation facts

Source-backed facts for citing this resource, derived directly from the registry — also available as plain text for AI assistants.

Canonical URL
https://heyclau.de/entry/mcp/windows-mcp-server
Source URLs
https://github.com/CursorTouch/Windows-MCP
Brand
Windows MCP
Safety notes
Windows MCP can interact with native Windows UI, open applications, control windows, simulate keyboard and mouse input, capture UI state, and automate browser content., Agents may interact with logged-in apps, desktop files, system dialogs, browsers, email clients, terminals, or enterprise software visible on the desktop., Use a test Windows account or VM for automation and require human approval before sending messages, submitting forms, changing settings, deleting files, or controlling sensitive applications., Background installation creates a per-user scheduled task and log files; review whether persistent background automation is appropriate before installing it.
Privacy notes
Window titles, UI text, file names, desktop state, browser content, application data, logs, and user input may be exposed to the MCP client and model., Automation logs can include private document names, customer data, credentials shown on screen, internal URLs, and enterprise application state., Browser DOM mode may expose page content from logged-in Chrome, Edge, or Firefox sessions.
Author
CursorTouch
Submitted by
oktofeesh1
Claim status
unclaimed
Last verified
2026-06-05

Decision playbook

Review trust signals before you adopt

Signals are present but mixed. Use the checklist below to confirm the source and operational safety for your environment.

Compare context
Selected

0

Current score

63

Baseline

—

Delta

No baseline selected

No major trust-signal divergence detected in the current selection.

Source and provenance checks

Needs review

Confirm ownership and provenance before trusting install instructions.

  • Source link availableRequired

    Open the canonical repository and verify ownership.

    Done

  • Source provenance statusRequired

    Marked as source-backed.

    Done

  • Metadata reviewed

    No reviewed flag detected in metadata.

    Pending

Safety and privacy checks

Complete

Validate risk disclosures before installation or API wiring.

  • Safety notes presentRequired

    Review the listed safety guidance before running commands.

    Done

  • Privacy notes presentRequired

    Review data handling notes before connecting accounts or secrets.

    Done

  • Trust level risk gateRequired

    Trust level does not block evaluation.

    Done

Package and install checks

Needs review

Check package metadata and artifact integrity signals.

  • Install payload available

    Install or copy payload is available for review.

    Done

  • Package verification flag

    No package verification flag provided.

    Pending

  • Checksum metadata

    No checksum provided for downloaded artifact.

    Pending

Compare-driven decision checks

Needs review

Use compare context to validate trade-offs before adoption.

  • Compare tray has multiple entries

    Add at least one more entry to compare trust differences.

    Pending

  • Baseline comparison available

    No baseline peer selected yet.

    Pending

  • Diverging trust signals identified

    No major trust-signal divergence found.

    Pending

Setup at a glance

CLI install

Copy-ready — paste the snippet to get started.

15 minutes

Install command

Provided

Config snippet

Provided

Copy snippet

Provided

Prerequisites

5 to clear

Platforms

4 listed

Install type

CLI install

Adoption plan

Balanced adoption plan

Current risk score 24/100. Use staged verification before broader rollout.

Risk 24

Pre-adoption checks

Validate source and review signals before any execution.

  • Confirm source provenanceRequired

    Source URL/provenance metadata is present.

    Done

  • Confirm metadata review state

    No review metadata found; increase manual validation.

    Pending

  • Verify install payload

    Install/config payload exists and can be inspected.

    Done

Security checks

Confirm safety, privacy, and package integrity signals.

  • Review safety notesRequired

    Safety notes are present.

    Done

  • Review privacy notesRequired

    Privacy notes are present.

    Done

  • Verify package integrity metadata

    No package verification/checksum metadata.

    Pending

Rollout

Adopt in controlled steps based on the selected plan.

  • Run in isolated sandbox firstRequired

    Use a constrained sandbox and observe behavior across multiple tasks.

    Pending

  • Roll out graduallyRequired

    Roll out to a small cohort before wider usage.

    Pending

  • Set monitoring and fallback

    Define rollback path and monitor errors after adoption.

    Pending

Evidence readiness

Evidence readiness matrix · balanced

Missing required evidence: Metadata review. Risk score 31.

Risk 31

Source provenance

Present

Source repository/provenance is listed.

Required in this preset

Metadata review

Missing

Review metadata is missing.

Required in this preset

Safety notes

Present

Safety notes are present.

Required in this preset

Privacy notes

Present

Privacy notes are present.

Optional in this preset

Package integrity

Missing

Package integrity metadata is missing.

Optional in this preset

Install payload

Present

Install payload is available.

Required in this preset

Required gaps: Metadata review

Decision timeline

Decision timeline · balanced

Blocking gaps: Check metadata review status. Risk 28.

Risk 28

triage

Confirm source provenanceRequired

Source/provenance metadata is available.

Done

triage

Check metadata review statusRequired

Review metadata is missing.

Pending

verify

Review safety notesRequired

Safety notes are available.

Done

verify

Review privacy notes

Privacy notes are available.

Done

verify

Validate package integrity metadata

Package integrity metadata is missing.

Pending

rollout

Verify install payload and commandsRequired

Install payload is available.

Done

Blockers: Check metadata review status

Prerequisite readiness

Prerequisite readiness

5 prerequisites to line up before setup.

0/5 ready
Install & runtime3General215 minutes

Safety & privacy surface

Safety & privacy surface

4 safety and 3 privacy notes across 3 risk areas. Review closely: credentials & tokens.

3 areas
  • SafetyGeneralWindows MCP can interact with native Windows UI, open applications, control windows, simulate keyboard and mouse input, capture UI state, and automate browser content.
  • SafetyLocal filesAgents may interact with logged-in apps, desktop files, system dialogs, browsers, email clients, terminals, or enterprise software visible on the desktop.
  • SafetyLocal filesUse a test Windows account or VM for automation and require human approval before sending messages, submitting forms, changing settings, deleting files, or controlling sensitive applications.
  • SafetyLocal filesBackground installation creates a per-user scheduled task and log files; review whether persistent background automation is appropriate before installing it.
  • PrivacyLocal filesWindow titles, UI text, file names, desktop state, browser content, application data, logs, and user input may be exposed to the MCP client and model.
  • PrivacyCredentials & tokensAutomation logs can include private document names, customer data, credentials shown on screen, internal URLs, and enterprise application state.
  • PrivacyCredentials & tokensBrowser DOM mode may expose page content from logged-in Chrome, Edge, or Firefox sessions.

Safety notes

  • Windows MCP can interact with native Windows UI, open applications, control windows, simulate keyboard and mouse input, capture UI state, and automate browser content.
  • Agents may interact with logged-in apps, desktop files, system dialogs, browsers, email clients, terminals, or enterprise software visible on the desktop.
  • Use a test Windows account or VM for automation and require human approval before sending messages, submitting forms, changing settings, deleting files, or controlling sensitive applications.
  • Background installation creates a per-user scheduled task and log files; review whether persistent background automation is appropriate before installing it.

Privacy notes

  • Window titles, UI text, file names, desktop state, browser content, application data, logs, and user input may be exposed to the MCP client and model.
  • Automation logs can include private document names, customer data, credentials shown on screen, internal URLs, and enterprise application state.
  • Browser DOM mode may expose page content from logged-in Chrome, Edge, or Firefox sessions.

Prerequisites

  • Windows 7, 8, 8.1, 10, or 11.
  • Python 3.13 or newer.
  • uv package manager installed.
  • English as the preferred Windows language, or the App tool disabled for other languages.
  • MCP client such as Claude Desktop, Perplexity Desktop, Gemini CLI, Qwen Code, or another compatible host.

Schema details

Install type
cli
Troubleshooting
No
Source repository stats
Scope
Source repo
Collection metadata
Estimated setup
15 minutes
Difficulty
intermediate
Full copyable content
{
  "mcpServers": {
    "windows-mcp": {
      "command": "uvx",
      "args": ["windows-mcp", "serve"]
    }
  }
}

About this resource

Content

Windows MCP is a computer-use MCP server for Windows. It bridges AI agents to the Windows operating system so they can inspect UI state, control apps and windows, simulate keyboard and mouse input, navigate files, and automate QA testing or desktop tasks.

The project also documents a browser DOM mode that focuses on web page content for Chrome, Edge, and Firefox automation.

Source Review

These sources were reviewed on 2026-06-05. Prefer the live repository for current client snippets, Windows requirements, transport options, persistent login-task behavior, and troubleshooting details.

Features

  • Native Windows UI automation for apps, windows, keyboard, and mouse actions.
  • UI state capture for agent reasoning.
  • File navigation and desktop workflow automation.
  • QA testing support for Windows applications.
  • Browser DOM mode for cleaner web automation in Chrome, Edge, and Firefox.
  • PyPI package with uvx windows-mcp serve setup.
  • Optional background task install for running the server at login.

Installation

Install uv, then configure your MCP client to run Windows MCP from PyPI:

{
  "mcpServers": {
    "windows-mcp": {
      "command": "uvx",
      "args": ["windows-mcp", "serve"]
    }
  }
}

Restart the client after saving the configuration. See the README for client-specific notes for Claude Desktop, Perplexity Desktop, Gemini CLI, Qwen Code, and other Windows MCP hosts.

Use Cases

  • Automate Windows desktop QA flows from an AI assistant.
  • Test application UI behavior by clicking, typing, and inspecting state.
  • Drive repetitive desktop workflows in a controlled Windows VM.
  • Use browser DOM mode for web automation that focuses on page content.
  • Let an agent open apps and gather UI state during troubleshooting.

Safety and Privacy

Windows MCP acts on the real Windows desktop. Use a test VM or dedicated account for automation, keep sensitive apps closed, and require confirmation before messages, purchases, account changes, system settings changes, or file deletion.

Treat UI state, browser DOM content, window titles, logs, file names, and screen text as sensitive. These can expose personal data, enterprise data, credentials, and private application state to the MCP client and model.

Duplicate Check

No CursorTouch/Windows-MCP entry or source URL was found in content/mcp. This entry is separate from browser-only automation servers and general terminal or filesystem MCP servers because it focuses on native Windows computer use.

#windows#desktop#automation#computer-use#testing

Source citations

Add this badge to your README

Show that Windows MCP Server is listed on HeyClaude. Paste this Markdown into your README — it renders the badge and links back to this page.

Listed on HeyClaude
[![Listed on HeyClaude](https://heyclau.de/badge/mcp/windows-mcp-server.svg)](https://heyclau.de/entry/mcp/windows-mcp-server)

How it compares

Windows MCP Server side by side with its closest alternative on trust, install, platform support, and disclosed safety notes — all from reviewed registry metadata.

1 trust signal differ across this comparison (Submitter).

Field
Windows MCP Server

Windows computer-use MCP server that lets AI agents interact with the Windows operating system through UI state, keyboard and mouse actions, application control, file navigation, QA testing, and browser DOM mode.

Open dossier
Hyperbrowser logoHyperbrowser MCP Server

Hyperbrowser's MCP server for AI agents that need hosted browser scraping, crawling, structured extraction, Bing search, persistent browser profiles, and browser-use, OpenAI CUA, or Claude computer-use browser agents.

Open dossier
Next steps
Open dossierAPI JSONOpen LLMMCP feedOpen sourceNewsletterClaim listing
Open dossierAPI JSONOpen LLMMCP feedOpen sourceNewsletterClaim listing
Trust
Review statusNot reviewedNot reviewed
Package trustPackage not verifiedPackage not verified
Source provenanceSource-backedSource-backed
SubmitterDiffersoktofeesh1—
Install riskReview firstReview first
Notes Safety ✓ Privacy ✓ Safety ✓ Privacy ✓
Brand—Hyperbrowser logoHyperbrowser
Categorymcpmcp
SourceSource-backedSource-backed
AuthorCursorTouchHyperbrowser
Added2026-06-052026-06-18
Platforms
Claude CodeCodexCursorClaude Desktop
Claude CodeClaude Desktop
Harness
Claude CodeCodexCursorClaude Desktop
Claude CodeClaude Desktop
Source repo——
Safety notes✓Windows MCP can interact with native Windows UI, open applications, control windows, simulate keyboard and mouse input, capture UI state, and automate browser content. Agents may interact with logged-in apps, desktop files, system dialogs, browsers, email clients, terminals, or enterprise software visible on the desktop. Use a test Windows account or VM for automation and require human approval before sending messages, submitting forms, changing settings, deleting files, or controlling sensitive applications. Background installation creates a per-user scheduled task and log files; review whether persistent background automation is appropriate before installing it.✓Hyperbrowser MCP can scrape pages, crawl linked pages, extract structured data, search with Bing, create persistent profiles, delete profiles, list profiles, and run browser automation agents. Browser-use, OpenAI CUA, and Claude computer-use tools can click, type, navigate, and interact with web apps. Keep human approval around authenticated, paid, destructive, or account-changing actions. Persistent browser profiles can retain session state. Delete profiles that should not be reused and avoid sharing profile identifiers in public logs or prompts. Scraping and crawling should respect site terms, robots expectations, rate limits, authentication boundaries, and internal data handling rules. Do not give the MCP server broad access to private web apps, admin consoles, payments, or production dashboards without a scoped task and review plan.
Privacy notes✓Window titles, UI text, file names, desktop state, browser content, application data, logs, and user input may be exposed to the MCP client and model. Automation logs can include private document names, customer data, credentials shown on screen, internal URLs, and enterprise application state. Browser DOM mode may expose page content from logged-in Chrome, Edge, or Firefox sessions.✓Webpage URLs, page content, screenshots or extracted data, search queries, browser actions, profile identifiers, and tool outputs may be processed by Hyperbrowser and the connected model provider. Authenticated browser sessions can expose cookies, account data, private documents, customer data, admin UI state, and generated artifacts to the MCP client and model provider. Keep `HYPERBROWSER_API_KEY`, cookies, one-time codes, session URLs, API responses, and extracted private data out of prompts, commits, issue comments, screenshots, and shared logs. For regulated or customer-sensitive data, review Hyperbrowser, MCP client, and model-provider retention policies before using browser automation.
Prerequisites
  • Windows 7, 8, 8.1, 10, or 11.
  • Python 3.13 or newer.
  • uv package manager installed.
  • English as the preferred Windows language, or the App tool disabled for other languages.
  • Node.js 18 or newer for the `hyperbrowser-mcp` npm package.
  • A Hyperbrowser account and `HYPERBROWSER_API_KEY` for hosted browser sessions.
  • An MCP client such as Claude Desktop, Claude Code, Cursor, Windsurf, or another client that can run stdio MCP servers.
  • Explicit approval for the URLs, websites, accounts, browser profiles, and data classes the agent may access.
Install
uvx windows-mcp serve
claude mcp add hyperbrowser -e HYPERBROWSER_API_KEY=your-api-key -- npx -y hyperbrowser-mcp
Config
{
  "mcpServers": {
    "windows-mcp": {
      "command": "uvx",
      "args": ["windows-mcp", "serve"]
    }
  }
}
{
  "mcpServers": {
    "hyperbrowser": {
      "command": "npx",
      "args": ["-y", "hyperbrowser-mcp"],
      "env": {
        "HYPERBROWSER_API_KEY": "YOUR-API-KEY"
      }
    }
  }
}
Citations
ClaimUnclaimedUnclaimed
Open in the interactive comparison tool

Related guides

Source-backed guides for putting this to work.

Computer Use From the Claude Code CLI for GUI QA

Drive desktop GUI QA from Claude Code with computer-use tools.

Added 1mo ago
guidesReview firstSource-backedReview first
Safety ✓ Privacy ✓by kiannidev

Scheduled Tasks In Claude Code Desktop

Schedule recurring Claude Code maintenance from Desktop with scoped tasks and run review.

Added 1mo ago
guidesReview firstSource-backedReview first
Safety ✓ Privacy ✓by kiannidev

Auto Mode Hard-Deny Policies For Safe Automation

Set autoMode.hard_deny rules to block risky actions in auto mode.

Added 1mo ago
guidesReview firstSource-backedReview first
Safety ✓ Privacy ✓by kiannidev

Signals

Loading live community signals…

More like this, weekly

A short, calm digest of reviewed Claude resources. Unsubscribe any time.