firecrawl — Claude Code skill: live web search and scraping
Wires the Firecrawl CLI into your agent so it can search the web, read pages, crawl sites and watch them for changes.
npx skills add firecrawl/cli --skill firecrawl -a claude-codeWhat firecrawl does
The skill grants the agent permission to run firecrawl and npx firecrawl-cli, then lets it pick the right command per task. search finds sources, scrape reads a known URL, map locates a subpage on a big site, crawl pulls a whole section such as /docs/, and monitor handles recurring checks. Pages come back as clean markdown sized for a model's context window.
Narrower commands cover specific ground: gov for US statutes and court opinions, research for scientific papers, developer for issues, merged PRs and library docs. Files already on disk go to firecrawl-parse instead. When a page needs clicks, form fills or a login, the flow is scrape first, then interact.
Run firecrawl --status before you start. It reports auth state, the concurrency limit and remaining credits. Nothing runs without authentication set up, so sort that out first. Search results fetched with --scrape already carry full page content, which means re-scraping those URLs just burns credits.
What it helps with
- gather sources on a topic and save them to files
- pull text from a JavaScript-rendered page
- locate one subpage inside a large site
- download an entire docs section in one pass
- set up recurring change alerts on a page
Example prompts for the agent
- Search the web for recent posts on vector databases and save the links
- Read the page at this URL and pull out the pricing table
- Set up a monitor on the releases page and tell me when it changes
How to install firecrawl
1Skills CLI (Vercel)
npx skills add firecrawl/cli --skill firecrawl -a claude-code2Manually: put the files in place
~/.claude/skills/firecrawl/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/firecrawl/
1Skills CLI (Vercel)
npx skills add firecrawl/cli --skill firecrawl -a codex2Manually: put the files in place
.agents/skills/firecrawl/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add firecrawl/cli --skill firecrawl -a cursor2Manually: put the files in place
.cursor/skills/firecrawl/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install firecrawl/cli firecrawl2Manually: put the files in place
.github/skills/firecrawl/SKILL.md1Skills CLI (Vercel)
npx skills add firecrawl/cli --skill firecrawl -a gemini-cli2Platform CLI
gemini skills install https://github.com/firecrawl/cli --path skills/firecrawlManually: put the files in place
.devin/skills/firecrawl/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add firecrawl/cli --skill firecrawl -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about firecrawl
How do I install firecrawl in Claude Code?
Run npx skills add firecrawl/cli --skill firecrawl -a claude-code. The skill lands at ~/.claude/skills/firecrawl/SKILL.md.
What does the firecrawl skill do?
It lets the agent run the Firecrawl CLI to search the web, scrape and extract pages, map and crawl sites, download a site to files, and watch pages for changes. Output arrives as markdown.
How is firecrawl different from firecrawl-parse?
firecrawl handles the live web only. Local files such as PDF, DOCX or XLSX route to firecrawl-parse.
