Hyreflow
Integrations

Firecrawl

Scrape, search, crawl, and extract web content as clean markdown. Key gotcha: blocked URLs (403/404) can still incur credits — never blind-retry.

Reach for Firecrawl whenever a recruiting/GTM step needs live web content: scrape a careers page, search the web with content in one call, map a site, or extract structured data with a schema. It is also the resolver of choice for LayoffSignal's Google News redirect links.

Managed credits or BYOK — flat 0.1 credits per scrape, search, or map call and per crawl page; extract is runtime-priced. Modifiers (PDF, JSON format, Enhanced Mode) stack.

Capabilities

ToolDoesCost
scrapeScrape a single page → markdown by default0.1 / request
searchWeb search + page content in one call0.1 / request
mapEnumerate a site's URLs before scraping0.1 / request
start_crawlFull-site async crawl (returns a job id)0.1 / page
extractStructured extraction via JSON schema or NL prompt (async)Runtime (usage-based)
get_crawlPoll crawl/job statusFree

Guidance

  • Pick by intent: single page → scrape; web search → search; discover then fetch → map then scrape the URLs you need; structured fields → extract.
  • Async ops (start_crawl, extract) return a job id immediately; the handler polls up to ~5 min, or poll get_crawl yourself for non-blocking use.
  • Cap your crawls: crawl defaults to a 10,000-page limit and preflights credits against it — always pass an explicit lower limit unless a full crawl is intended.
  • Don't blind-retry blocked URLs: Firecrawl can charge even when the target returns 403/404 — inspect metadata.statusCode first.
  • Use markdown for LLM consumption (default), html for raw DOM, links for hyperlinks, screenshot when layout matters.
hyreflow tools execute firecrawl scrape \
  --payload '{"url":"https://example.com/careers","formats":["markdown"]}'

Hand a LayoffSignal redirector link to Firecrawl — it follows the Google News redirect and returns the publisher article body.

On this page