Fetcher MCP
by jae-jae
Fetcher MCP is an MCP server that fetches web page content using the Playwright headless browser, enabling JavaScript support and intelligent content extraction. It supports both HTML and Markdown output formats and allows for parallel processing of multiple URLs.
Last updated: N/A
What is Fetcher MCP?
Fetcher MCP is a server that utilizes Playwright to fetch web page content, supporting JavaScript execution and intelligent content extraction. It's designed to overcome the limitations of traditional web scrapers by handling dynamic web content.
How to use Fetcher MCP?
You can run Fetcher MCP directly with npx -y fetcher-mcp. First-time users need to install the required browser with npx playwright install chromium. Debug mode can be enabled with the --debug option. It can be configured in Claude Desktop by adding a configuration in the claude_desktop_config.json file.
Key features of Fetcher MCP
JavaScript Support
Intelligent Content Extraction
Flexible Output Format (HTML and Markdown)
Parallel Processing
Resource Optimization
Robust Error Handling
Configurable Parameters
Use cases of Fetcher MCP
Extracting content from dynamic websites
Batch processing of multiple URLs
Integrating web scraping into downstream applications
Automated data collection
Fetching content for AI models
FAQ from Fetcher MCP
How do I handle websites with anti-crawler mechanisms?
How do I handle websites with anti-crawler mechanisms?
Use the prompt 'Please wait for the page to fully load' to enable waitForNavigation: true and increase the timeout duration with 'Please set the page loading timeout to 60 seconds'.
How do I preserve the original HTML structure?
How do I preserve the original HTML structure?
Use the prompt 'Please preserve the original HTML content' which sets extractContent: false and returnHtml: true.
How do I fetch the complete webpage content?
How do I fetch the complete webpage content?
Use the prompt 'Please fetch the complete webpage content instead of just the main content' which sets extractContent: false.
How do I return content in HTML format?
How do I return content in HTML format?
Use the prompt 'Please return the content in HTML format' which sets returnHtml: true.
How do I enable debug mode?
How do I enable debug mode?
Use the prompt 'Please enable debug mode for this fetch operation' which sets debug: true. Alternatively, run the server with the --debug flag.