Web Fetch

web_fetch fetches web page content and returns readable text. It's more of a "page-fetching tool" rather than a "generic API client".

What it does

• only accepts http:// or https://
• follows redirects by default
• extracts readable text from HTML by default
• returns status code, headers, final URL, and content
• blocks private network and loopback addresses by default

Parameters

Besides the required url, these commonly used parameters are supported:

Parameter	Description
method	defaults to GET, also supports POST, PUT, PATCH, DELETE, HEAD
headers	JSON object of request headers
params	JSON object of query parameters
body	request body
timeout	timeout in seconds
max_bytes	maximum returned bytes
follow_redirects	whether to follow redirects
extract_text	extract text from HTML
user_agent	custom User-Agent

Config

tools:
  web_fetch:
    enabled: true
    timeout: 20
    max_bytes: 200000
    follow_redirects: true
    user_agent: "MushroomAgent web_fetch/1.0"
    allow_private_networks: false

When to use it

• fetch public page content
• pull documentation, announcements, static HTML pages
• when you don't need complex authentication or API semantics

For structured HTTP APIs, internal services, or precise API path control, use API Request.