$ loading_
Control a computer automatically using mouse, keyboard, OCR, and window management.
This is an open-source MIT-licensed local computer-control MCP with no declared secrets or remote endpoints, and no clear high-risk red flags are evident from the provided material. Its core capabilities involve local input control and screen/window/OCR access, which warrant caution as inherent tool capabilities.
The material explicitly states that no keys or environment variables are required, and no API tokens, account credentials, or external authentication requirements are shown, so credential exposure appears limited.
The material declares no remote endpoints, and the description only mentions local mouse, keyboard, OCR, and screen/window management; there is no evidence of user data being sent to external services.
System checks indicate the tool has code-execution capability, and its functionality inherently allows control of local mouse, keyboard, and windows, meaning it can perform high-impact actions on the local machine on the user's behalf. This kind of local control is a typical high-privilege MCP capability and should be granted cautiously.
The description includes OCR and screen/window management, indicating access to screen contents and UI context, which may indirectly expose sensitive information. No README details are provided to further constrain readable/writable scope, so a relatively broad local visible-data access surface should be assumed.
Positive factors include being open source, auditable, and MIT-licensed; however, the source is a third-party registry, community adoption is only 0 stars, maintenance status is unknown, and the provided material lacks README detail, so supply-chain trust is not strong enough to rate as fully low risk.
Copy the install command and let the AI configure it · recommended for beginners
No copy-paste install info for "computer-control-mcp" yet — see the docs or source repo.
Please control the computer to open the browser, log in to the admin system, go to the orders page, filter unprocessed orders from today, and copy the order IDs into a text file; if page elements cannot be read directly, use OCR as needed.
The tool completes the web actions and produces a text file with order IDs or a result summary.
Please launch the local desktop app, click through the Login, Settings, and Export pages in order, record whether each step displays correctly, and save screenshots if any error dialog appears.
Returns a step-by-step execution report with error screenshots for basic UI regression testing.
Please switch through the currently open windows one by one, extract each window's title and key figures, and organize them into a table; use OCR for any content that cannot be copied directly.
Outputs a structured table summarizing the key information from each window.
Let AI control your computer for web, app, and desktop tasks.
Let AI control macOS apps, mouse, and keyboard to automate tasks.
Control desktop workflows locally with OCR and visible safety overlays.
Monitor and control your computer with real-time system, file, code, and web tools.
Expose raw desktop, shell, and screenshot control to AI clients.
Lets AI locally see and control your desktop for automation tasks.