Frequently Asked Questions (FAQ) about Puppeteer Vision MCP Server

Q: What is the Puppeteer vision MCP Server? A: The Puppeteer vision MCP Server is a tool for scraping webpages and converting them to markdown format using Puppeteer, Readability, and Turndown. It features AI-driven interaction capabilities to handle cookies, captchas, and other interactive elements automatically.

Q: What is MCP? A: MCP stands for Model Context Protocol. It is an open protocol that standardizes how applications provide context to Large Language Models (LLMs). The MCP server acts as a bridge, allowing AI models to access and interact with external data sources and tools.

Q: How does the AI-driven interaction work? A: The system uses vision-capable AI models to analyze screenshots of web pages and decide on actions like clicking, typing, or scrolling to bypass overlays and consent forms. This process repeats up to maxInteractionAttempts.

Q: What environment variables are required to run the server? A: The required environment variable is OPENAI_API_KEY. Optional variables include VISION_MODEL, API_BASE_URL, USE_SSE, PORT, and DISABLE_HEADLESS.

Q: What are the communication modes supported by the server? A: The server supports two communication modes: stdio (default) and SSE mode. Stdio communicates via standard input/output, while SSE mode communicates via Server-Sent Events over HTTP.

Q: How do I use the scrape-webpage tool? A: The scrape-webpage tool requires the url parameter (the URL of the webpage to scrape). Optional parameters include autoInteract, maxInteractionAttempts, and waitForNetworkIdle.

Q: What is the response format of the scrape-webpage tool? A: The tool returns its result in a structured format with content (an array containing a single text object with the raw markdown) and metadata (status message, success boolean, and content size).

Q: Can I run the browser in visible mode? A: Yes, you can set the DISABLE_HEADLESS environment variable to true to run the browser in visible mode.

Q: How do I install the server for development? A: Clone the repository, install dependencies using npm install, build the project using npm run build, set up your environment variables, and run the server using npm start or npm run dev.

Q: What is UBOS, and how does the MCP Server integrate with it? A: UBOS is a full-stack AI Agent Development Platform focused on bringing AI Agents to every business department. The Puppeteer vision MCP Server can be integrated as an asset within the UBOS platform, allowing AI Agents to access and scrape web content seamlessly.

Q: Where can I find the Puppeteer vision MCP Server on UBOS? A: You can find it on the UBOS Asset Marketplace.

Q: Can I customize the behavior of the scraper? A: Yes, you can modify the behavior of the scraper by editing files such as src/ai/vision-analyzer.ts, src/ai/page-interactions.ts, src/scrapers/webpage-scraper.ts, and src/utils/markdown-formatters.ts.