Firecrawl - BoostGPT Documentation

Scrape, crawl and map a site into clean text. Point it at a competitor to see how they explain what you also do.

Connecting

In the dashboard, open Connectors, find Firecrawl and press Connect. The key you enter is checked against Firecrawl before anything is saved, so a key that is refused never becomes a connection that fails later.

What you will need

Field	What it is
API Key	Your Firecrawl API key

Getting the credential

Create an API key at firecrawl.dev — it starts with fc-
Paste it into the connector

Notes

scrape renders JavaScript, so it works on pages a plain fetch returns empty.
crawl is asynchronous and charged per page. It returns an id; poll get_crawl until completed. Set limit deliberately — the default here is 25, not Firecrawl’s maximum.
Use map first when you only need to know what exists on a site; it costs a fraction of a crawl.

What an agent can do with it

Tool	What it does
`scrape`	Fetch one page and return it as clean markdown, with navigation and footers stripped. Handles JavaScript-rendered pages, which a plain fetch does not.
`crawl`	Start a crawl of a site and its linked pages. Asynchronous: returns an id — poll get_crawl until the status is ‘completed’.
`get_crawl`	Poll a crawl started by crawl. Status is ‘scraping’, ‘completed’ or ‘failed’.
`map`	List the URLs on a site without fetching them. Far cheaper than crawling.
`search`	Search the web and return each result already scraped as markdown.

Useful resources

Firecrawl documentation