Capture screenshots, console errors, network logs, and Lighthouse audits — and feed them directly to your AI coding assistant via MCP.
If you want to debug a website or automate browser tasks, you normally can't do it from AI apps like Cursor or Claude Code because they don't have access to a web browser. RapidTriageME bridges that gap with a Chrome Extension + MCP Server that captures browser data and feeds it directly to your AI assistant. Everything stays local on your machine.
Speed up and automate your debugging workflows
Your AI assistant captures browser errors, analyzes root causes with confidence scores, and suggests code fixes — no more copy-pasting stack traces.
Let your AI editor test your code end-to-end. Validate UI elements, check network responses, and ensure your application works across scenarios.
Run Lighthouse audits for performance, accessibility, SEO, and best practices on any URL. Results flow directly to your AI assistant for analysis.
Debug with speed, privacy, and convenience
Browser data captured locally via the Chrome extension. No proxy servers or network round-trips.
Browser activity stays on your device. Data is only shared with your AI assistant via MCP.
Uses your existing browser session. Stays logged into all your services — no separate credentials.
Works with Cursor, Claude Code, VS Code, Windsurf, Zed, JetBrains, Neovim, and more.
Get started in three simple steps
Add the RapidTriageME extension to your browser. It captures screenshots, console logs, and network data.
Add the RapidTriageME MCP server to your AI application. One JSON config connects your IDE to the browser.
Your AI assistant can now see your browser. Ask it to capture errors, run audits, or triage bugs.
Everything you need to debug, test, and audit your web app
Full-page and viewport captures
Real-time errors, warnings, messages
XHR/fetch requests and responses
Select and inspect elements
LCP, FID, CLS, TTI, Speed Index
WCAG contrast, ARIA, keyboard nav
Meta tags, structured data, crawlability
Security headers, HTTPS, standards
Execute code in browser context
Go to URLs, back, forward
Root cause analysis + confidence scores
Run all audits sequentially
Start free. Upgrade as your team grows.
RapidTriageME connects your AI coding assistant (Cursor, Claude Code, VS Code, Windsurf, etc.) to your browser via the Model Context Protocol (MCP). It captures screenshots, console errors, network logs, and runs Lighthouse audits — so your AI can debug your web app without you copy-pasting anything.
Install the Chrome Extension and add the MCP server to your IDE config. Your IDE connects to the MCP server, which communicates with the extension to capture real-time browser data. The MCP protocol is an open standard supported by all major AI coding tools.
Any IDE that supports the Model Context Protocol (MCP), including Cursor, Claude Code, VS Code (with Continue), Windsurf, Zed, JetBrains IDEs, Neovim, and more. If your editor supports MCP, it works with RapidTriageME.
Yes. The remote SSE server can capture screenshots and run Lighthouse audits on any public URL without the extension. Connect via the SSE transport in your MCP config.
Browser data is captured locally by the Chrome extension and shared only with your AI assistant via the MCP protocol. We never collect or store your browsing activity. Enterprise customers get dedicated isolated realms with SSO support.
Sign up at rapidtriage.me, go to Settings, and create an API key. Use this token in your IDE's MCP configuration to authenticate with the server.
Install RapidTriageME and let your AI assistant debug your web app.
Add to Chromechrome://extensionsChrome Web Store listing coming soon. For now, install as an unpacked extension.