Firecrawl

Saves every page of a site as markdown, HTML, or screenshots — works offline and lets you search, edit, or archive docs without relying on the live site.

Runs a web search and pulls the full text from each result—not just snippets. Read articles, compare sources, or find recent news without leaving Claude.

Navigates dynamic sites—pricing pages, product catalogs, directories—and pulls data into clean JSON matching a schema you define. Handles JavaScript-heavy sites.

Points at a URL and extracts every page under that path—docs sections, help centers, competitor sites—into clean markdown with preserved structure and links.

Controls a live browser to fill forms, click buttons, and navigate multi-step flows. Tell it what you need in plain language and it extracts the data after interacting.

Pulls the readable content from a URL — including JavaScript-heavy sites — and returns it as clean markdown you can pipe into other skills or analyze.

Sets up a watch on any URL and triggers a notification (webhook or email) the moment content updates—no polling scripts needed. Catches pricing shifts, new postings, docs updates.

Crawls a domain and indexes every URL, then searches for pages matching a keyword or phrase. Shows site structure and pinpoints exactly where content lives.

Searches, scrapes, and extracts structured data from any URL — articles, documentation, product pages, forums. Returns clean markdown and JSON for downstream work.

Converts PDFs, Word docs, spreadsheets, and HTML files into structured markdown—removing junk, preserving tables and links, saving the result to your computer.

Integrates Firecrawl's search endpoint to query the web, rank results by relevance, and optionally extract structured data from matching pages — no manual URL hunting.

Walks through fetching an API key, installing the SDK, and setting up environment variables so your code can crawl and scrape sites without the boilerplate.

Integrates Firecrawl into your product code so a single URL returns markdown, screenshots, links, and structured metadata—ready to feed into another tool or workflow.

Integrates Firecrawl's `/interact` endpoint to click, fill forms, paginate, and authenticate—extracting data from sites that need browser actions, not just static HTML parsing.