Firecrawl

The Arcade Firecrawl MCP Server provides a pre-built set of tools for interacting with websites. These tools make it easy to build agents and AI apps that can:

• Scrape web pages
• Crawl websites
• Map website structures
• Retrieve crawl status and data
• Cancel ongoing crawls

Available Tools

These tools are currently available in the Arcade Firecrawl MCP Sever.

Firecrawl.ScrapeUrl

Scrape a URL and return data in specified formats.

Auth:

• Environment Variables Required:
  • FIRECRAWL_API_KEY: Your Firecrawl  API key.

Parameters

• url (string, required) The URL to scrape.
• formats (enum (Formats), optional) The format of the scraped web page. Defaults to Formats.MARKDOWN.
• only_main_content (bool, optional) Only return the main content of the page. Defaults to True.
• include_tags (list, optional) List of tags to include in the output.
• exclude_tags (list, optional) List of tags to exclude from the output.
• wait_for (int, optional) Delay in milliseconds before fetching content. Defaults to 10.
• timeout (int, optional) Timeout in milliseconds for the request. Defaults to 30000.

Firecrawl.CrawlWebsite

Crawl a website and return crawl status and data.

Auth:

• Environment Variables Required:
  • FIRECRAWL_API_KEY: Your Firecrawl  API key.

Parameters

• url (string, required) The URL to crawl.
• exclude_paths (list, optional) URL patterns to exclude from the crawl.
• include_paths (list, optional) URL patterns to include in the crawl.
• max_depth (int, required) Maximum depth to crawl. Defaults to 2.
• ignore_sitemap (bool, required) Ignore the website sitemap. Defaults to True.
• limit (int, required) Limit the number of pages to crawl. Defaults to 10.
• allow_backward_links (bool, required) Enable navigation to previously linked pages. Defaults to False.
• allow_external_links (bool, required) Allow following links to external websites. Defaults to False.
• webhook (string, optional) URL to send a POST request when the crawl is started, updated, and completed.
• async_crawl (bool, required) Run the crawl asynchronously. Defaults to True.

Firecrawl.GetCrawlStatus

Retrieve the status of a crawl job.

Auth:

• Environment Variables Required:
  • FIRECRAWL_API_KEY: Your Firecrawl  API key.

Parameters

• crawl_id (string, required) The ID of the crawl job.

Firecrawl.GetCrawlData

Retrieve data from a completed crawl job.

Auth:

• Environment Variables Required:
  • FIRECRAWL_API_KEY: Your Firecrawl  API key.

Parameters

• crawl_id (string, required) The ID of the crawl job.

Firecrawl.CancelCrawl

Cancel an ongoing crawl job.

Auth:

• Environment Variables Required:
  • FIRECRAWL_API_KEY: Your Firecrawl  API key.

Parameters

• crawl_id (string, required) The ID of the asynchronous crawl job to cancel.

Firecrawl.MapWebsite

Map a website from a single URL to a map of the entire website.

Auth:

• Environment Variables Required:
  • FIRECRAWL_API_KEY: Your Firecrawl  API key.

Parameters

• url (string, required) The base URL to start crawling from.
• search (string, optional) Search query to use for mapping.
• ignore_sitemap (bool, required) Ignore the website sitemap. Defaults to True.
• include_subdomains (bool, required) Include subdomains of the website. Defaults to False.
• limit (int, required) Maximum number of links to return. Defaults to 5000.

Auth

The Arcade Web MCP Sever uses Firecrawl  to scrape, crawl, and map websites.

Global Environment Variables:

• FIRECRAWL_API_KEY: Your Firecrawl  API key.

pip install arcade_firecrawl