ipIterPrompt
MCP serverWeb & SearchVerified Last verified 2026-07-24

Fetch MCP Server

by Model Context Protocol

A reference MCP server that retrieves a web page and converts it to markdown for efficient LLM consumption. It supports chunked reading via a start index (so the model can page through long content), raw HTML access, a customizable user-agent, proxy support, and robots.txt handling.

Works with

The Model Context Protocol (MCP).

When to use it

  • Letting an agent read a specific documentation page or article
  • Pulling a known URL's content into context without a full browser
  • Lightweight, key-free web reading

When not to

  • Pages that require interaction or login — use a browser-automation server instead
  • Untrusted internal URLs — it can reach local/internal addresses, so restrict use

What it does

  • Fetch any URL and extract its content as markdown
  • Read long pages in chunks using a start index
  • Optionally return raw HTML and respect robots.txt

How to install

Setup steps with copy-paste config. Sources are cited under each method.

uvx (Python)

Requires: Python with uv/uvx

  1. Add to your MCP config

    No API key needed. Runs the Python server directly with uvx.

    json
    {
      "mcpServers": {
        "fetch": {
          "command": "uvx",
          "args": ["mcp-server-fetch"]
        }
      }
    }
  2. Ask the agent to read a URL

    Give it a link and it will fetch and summarize the page content.

How to prompt it

Copy-paste prompts written to exercise this plugin, with what to expect.

Summarize a docs page

You want the gist of a long documentation page.

Fetch https://example.com/docs/getting-started and give me the setup steps as a short checklist.

What to expect: The agent retrieves the page, converts it to markdown, and returns a concise checklist grounded in the real content.

  • For very long pages, ask it to continue reading from where it left off.

Tips for getting the most out of it

  • Great paired with a search server: search to find URLs, fetch to read them.
  • Be mindful that it can reach internal addresses; don't point it at untrusted links.

Prompts to pair with it

Real prompts from the IterPrompt library you can copy or run.

Modify Front-End Webpage with Codex and Image Input

Turn any AI chat into a modify front-end webpage with codex and image input with this community persona prompt.

189 copiesOpen ↗

One-Page Website Coder

A community prompt for one-page website coder, imported from LLM-Prompt-Library (MIT).

3,459 copiesOpen ↗

Policy Enforcer Agent

Cedar policy author and reviewer for Claude Code tool calls. Writes, audits, and explains Cedar policies that govern Bash, Edit, Write, WebFetch, and other tools. Use when you need

4,676 copiesOpen ↗

One-Page PRD Writer

Turn a rough feature idea into a crisp one-page PRD with open questions.

4,870 copiesOpen ↗

Landing Page Copy, Section by Section

Write conversion-focused landing page copy from a positioning statement.

3,540 copiesOpen ↗
4.6

SEO Page Builder

Every page your agent ships comes out technically SEO-complete.

2,170 copiesOpen ↗
See more prompts in the library

Frequently asked questions

Does it obey robots.txt?+

By default it respects robots.txt for model-initiated requests; that can be disabled via a flag. It has no search of its own — you supply the URL.

Pairs well with

Alternatives to Fetch MCP Server

Sources & verification

The facts on this page and where they come from. Anything unverified is labeled — we'd rather say "unknown" than guess.

Install steps, pricing, and availability change often. Always confirm on the official page before relying on setup details.