MCP server

Firecrawl web scraping

Name: Firecrawl web scraping
Author: Firecrawl

by Firecrawl @firecrawl

Scrape, crawl and search the web, turning pages into clean text for the agent.

Listed by Terminal Deck — Firecrawl has not claimed this profile.

What it is: MCP server — A server that gives an agent new tools.
What it needs: Needs Node.js Needs a key you supply Needs an account somewhere
On your machine: Runs a program on this machine
Who made it: Firecrawl @firecrawl
Price: Free, needs an account

Source on github.com Website

Needs a Firecrawl account and API key; Firecrawl sets its own limits and prices.

scraping crawl web search extract

How to install it

Installing from Commons is not in version 0.15.0, the version you can download today — the Community shelf arrives in the next release. Until then this page is the record of what this is, what it would put on your machine, and where it comes from.

Open Terminal Deck, go to Store, open the Community shelf, search for Firecrawl web scraping and press Install.
It is added as an MCP server your agents can call.

The command it runs, built by Terminal Deck and not by the publisher:

npx -y firecrawl-mcp

It asks you for one thing first, and keeps it in your machine's own keychain:

FIRECRAWL_API_KEY — Firecrawl API key (From firecrawl.dev, under API keys)

The facts

Publisher: Firecrawl
Kind: MCP server
Version: 3.24.0
Licence: MIT
Shelf: Searching and reading the web
Repository: firecrawl/firecrawl-mcp-server
Commit: 8c93c5617ed2674e30e8bf828699a59641a3d534
Last updated: 28 August 2026

Ratings

Not enough ratings yet. A rating appears at five, so one vote cannot speak for a listing.

For AI readers

/@firecrawl/firecrawl/llms.txt — the facts on this page as plain text. Firecrawl has written no description of their own.

What this file is for

An assistant that lands on this page has to work out what this is from headings and a one-line summary, and it will guess. So every listing is also served as one plain-text file, at the same address with /llms.txt on the end. A publisher can put their own words in it by naming a text or markdown file in their terminaldeck.json, or by typing one when they publish.

The whole catalogue is served the same way: every listing as one line, or every listing in full with each publisher's own description. Everything Firecrawl publishes is at /@firecrawl/llms.txt.

Discussion

Ask Firecrawl and everybody else using this.

Terminal Deck Moderator 3 days ago

This one needs a Firecrawl account and your own key. The key is kept on your machine — Terminal Deck never sees it, and there is no field on this listing that could send it anywhere.

More by Firecrawl → All of Commons → Ask the community →