fetchLABSCO SUMMARY
One fetch, whatever shape you need.
The conversion step disappears: ask for a page as Markdown or an API response as readable JSON and that is what comes back, sanitized on the way.
What it is
A Node server that retrieves web content and converts it. HTML, CSV, XML and JSON sources are parsed with Cheerio, Turndown, Fast-XML-Parser and CSVtoJSON, sanitized, and returned as whichever format you named.
What you get
- fetch retrieves a URL and picks the output format automatically
- fetch-markdown, fetch-text, fetch-json and fetch-html pin the output to one format regardless of what the source was
- extractText strips a page down to its content; prettyPrint formats JSON for reading
- CSV becomes a table, XML becomes JSON, and HTML carries its metadata across
- Content is sanitized with SanitizeHTML on the way back
- A history of recently fetched URLs
Requirements
Node.js 16.x or higher, cloned and built from the repository: npm install, npm run build, npm start. No account or key.
Setup effort
Build from source — git clone https://github.com/nathanonn/mcp-url-fetcher.git
USE CASES
01
Reading a long page without the markup
The article comes back as Markdown or as stripped text, which is both easier to read and far smaller in context.
02
Pulling an API response into the conversation
JSON arrives formatted rather than as one line, so the field you are looking for is findable.
03
Getting a CSV as something you can reason about
A raw CSV file is converted to a table on the way in, without a download and a spreadsheet in between.
04
Keeping the raw HTML when you need the structure
Sometimes the markup is the point — fetch-html returns it, sanitized, with its metadata intact.
05
Reducing a page to plain text
Plain text for the cases where formatting is noise, such as feeding a page into a summary.
7 tools
Listed in project docs. Your client may show a different list.
fetch-jsonfetch-htmlfetch-markdownfetch-textweb-searchdeep-researchBEFORE YOU INSTALL
What to know first
Tools you get
7 tools named in the project's documentation.
Requires
Node.js 16 or later
Where it runs
The documented client runs the server on your machine.
INSTALLATION
Connect your client
You clone and build this one yourself.
Server setup
TERMINAL
git clone https://github.com/nathanonn/mcp-url-fetcher.git
cd mcp-url-fetcher
npm install
npm run buildMCP Clients
Claude Desktop
CLAUDE DESKTOP · JSON
{
"mcpServers": {
"url-converter": {
"command": "node",
"args": [
"<YOUR_PATH_TO_MCP_URL_FETCHER>/build/index.js"
]
}
}
}
