$ loading_
Capture local web pages and UI elements with metadata for precise AI analysis.
This MCP tool is described as taking screenshots of local web pages and capturing user-selected UI elements, with no required secrets and no declared remote endpoints; no clear high-risk red flags are evident from the materials. Its core functionality inherently involves local code execution and access to local page/UI content, which warrants caution rather than a high-risk rating, especially given the open-source MIT-licensed source.
The materials explicitly state that no keys or environment variables are required. No API token, account credential, or other long-lived secret is requested, so credential exposure risk appears low.
No remote endpoint is declared, and the description focuses on local webpage screenshots and element capture. The provided materials do not indicate transmission of user data to external services, so no explicit egress path is evident.
The system checks indicate that this tool executes code. Given its screenshot and element-capture functionality, it likely starts local processes or invokes browser/automation components to access pages and UI elements. This is a normal MCP capability and merits caution about local execution scope, but not a high-risk rating by itself.
Its functionality requires access to local webpage render output, screenshot contents, and metadata for user-selected UI elements, which may expose sensitive information visible on the page. The materials do not show requests for broad filesystem permissions beyond the stated purpose, but access to screen/page content itself deserves caution.
The project has a public GitHub repository and an MIT open-source license, which are meaningful risk-reducing factors because the code is auditable. However, it comes from a third-party registry, has 0 stars, and its maintenance status is unknown, so overall trust and maturity remain limited; reviewing the code and dependencies before use is advisable.
Copy the install command and let the AI configure it · recommended for beginners
No copy-paste install info for "Peek" yet — see the docs or source repo.
Use Peek to capture the checkout page in my local development environment and select the "Pay Now" button. Analyze its visibility, hierarchy, size, and surrounding elements for usability issues.
A page screenshot, button metadata, and an analysis of usability issues with improvement suggestions.
Use Peek to open the local web page and capture the user profile card. Extract element details for the card title, avatar area, and action buttons, then check whether they match the design spec.
A screenshot, metadata for the selected elements, and a summary of differences from the design spec.
Use Peek to capture the local dashboard page and select the left navigation and top filter area. Organize the screenshot and element metadata for a later AI redesign recommendation.
A reusable page screenshot, metadata for key regions, and structured context for later analysis.
Capture macOS screenshots and analyze interface content with visual question answering.
Expose local browser session recordings to AI agents for debugging and analysis.
Capture and automate macOS UI actions with Peekaboo CLI for scripting and testing.
Lets AI capture localhost screenshots to verify UI changes visually.
Lets AI understand and operate web interfaces through unified UI scene graphs.
Capture any URL or HTML as high-quality images for preview and testing.