Browser Tools MCP provides web browsing capabilities for AI assistants through the @agentdeskai/browser-tools-mcp package. It runs as a Node.js application in a Docker container, configurable via smithery.yaml to adjust port settings and log levels. The implementation uses a non-root user for security and supports both standard and Server-Sent Events (SSE) communication modes. This tool enables AI models to interact with web content, making it valuable for workflows requiring web scraping, content analysis, or automated browsing tasks.
暂无评论. 成为第一个评论的人！
登录以参与讨论
Highlight AIConsole output, filterable by keyword with paging.
Error-level output and uncaught exceptions.
XHR/fetch requests with status, timing and bodies.
Only failed and 4xx/5xx requests.
The element selected in the Elements panel.
Which page the browser is currently on.
Whether the extension is connected, and capture counts.
Every tab with DevTools open, and the id to address it by.
Screenshot returned as an image, plus its file path.
Reloads the inspected tab.
localStorage, sessionStorage and cookies (values gated).
Clears captured telemetry before a clean reproduction.
Lighthouse accessibility audit.
Lighthouse performance audit with Core Web Vitals.
Lighthouse SEO audit.
Lighthouse best-practices audit.
