fetch-mcp

Fetch MCP Server

fetch-mcp retrieves web content through a Playwright browser for MCP client users working with JavaScript-driven pages. It supports individual and parallel URL requests, with main-content extraction and Markdown or HTML output.

Server facts

fetch-mcp

Project

Author
jae-jae
Website
1team.top
License
MIT

Tech

Runtime / language
node
Transports
stdio
Remote available
No
Authentication
Not specified

Activity

GitHub stars
1,088
Last commit
Data as of

Tags

  • Playwright
  • Web Scraping
  • fetch

Overview

Dynamic web content is the focus of fetch-mcp, which uses Playwright to execute page JavaScript before returning content to an MCP client. The fetch_url tool handles an individual address, while fetch_urls processes multiple addresses concurrently in browser tabs and separates their results in the combined response. Readability-based extraction selects the main text and removes surrounding material such as navigation and advertising. Users can instead request the full page content, choose HTML rather than Markdown, or cap the length of the returned content.

Retrieval settings cover loading timeouts, navigation completion conditions, and an additional navigation wait for pages involving redirects or verification steps. Images, stylesheets, fonts, and media are blocked by default, with an option to change that behavior. For troubleshooting or authenticated access, debug mode exposes the browser window so a user can sign in manually before retrieval continues. The browser_install tool supplies the Chromium binary, with options to include system dependencies or force reinstallation. The project runs on Node and carries an MIT license.

fetch-mcp setup notes

  • fetch-mcp records stdio only; confirm a network transport upstream before planning remote access.

Related MCP servers