Skip to directory

MCP Science ? Web Fetch

Path Integral Institute (community maintained)

stdio2+ documented tools18

About MCP Science ? Web Fetch

The web-fetch server from the MCP Science project fetches and processes HTML, PDF, and plain-text web content for assistant research.

What you can use it for

  • web content retrieval
  • HTML and PDF research
Connection and setup
Hosting
Local process
Transport
stdio
Endpoint
Local process; no hosted endpoint documented
Authentication
No credential is documented for the web-fetch sub-server; runtime authentication status has not been tested.

Requirements

  • UV/uvx runtime; repository docs do not specify a minimum version.
  • Network access to URLs the user asks the server to fetch.

Setup steps

  1. Use an MCP client with stdio support and install UV/uvx using Astral's documented installer if it is not already present.
  2. Add a local server with command uvx and args ["mcp-science", "web-fetch"].
  3. Restart or reconnect the client, then check its MCP view for successful initialize and tools/list.
  4. For a functionality check, use fetch-web on a public, non-sensitive URL only after the user approves the target.
Tools 2+

The project README lists these two tools for the web-fetch sub-server. No runtime tools/list was performed.

  • fetch-web

    Fetches and processes HTML, PDF, and plain-text web content.

    01
  • read-image-url

    Reads an image from a URL.

    02

This server’s available tools can vary by account, permissions, or configuration. The list shown is not its full runtime catalog.

Limitations and notes
  • MCP Science is a community-maintained multi-server project; the web-fetch sub-server is a deployment variant, not the entire MCP Science project API.
  • Runtime initialize, tools/list, and function calls have not been tested in an OS-isolated environment.
Original sources
  • MCP.Directory public profile

    Used for discovery and identifying the web-fetch server variant; directory setup claims were checked against the linked repository.

  • Community-maintained repository README

    Documents the uvx mcp-science web-fetch command and describes web-fetch behavior. The README separately identifies other sub-servers with their own credentials.

  • Developer organization logo in linked repository

    The downloaded image visibly displays the Path Integral developer mark. It was checked on light/dark backgrounds and selected as organization-logo tier B; it is not represented as an official MCP product logo.

  • maintainer repository reference

    Repository documents the web-fetch MCP variant and its fetch-web and read-image-url tools; separate subservers have different prerequisites.

Documentation checked 2026-10-28T00:00:00Z. Connection has not been runtime tested.

Similar tools