firecrawl-agent — Claude Code skill: extraction and setup
This skill navigates websites on its own and extracts structured data. Use it for multi-page catalogs, pricing data, and listings.
npx skills add firecrawl/cli --skill firecrawl-agent -a claude-codeWhat firecrawl-agent does
firecrawl-agent gives an agent a navigation and extraction task. It can save JSON with --wait --json -o, limit the run to selected pages with --urls, or shape the result with --schema. For a simple single-page extraction, use scrape instead: agent runs consume more credits.
Before autonomous extraction of records or listings, the skill checks whether an Alexandria provider already covers the request. It can search with firecrawl search, inspect a contract with firecrawl list, and run the matching capability through firecrawl scrape --alexandria. When no suitable provider exists, the agent navigates the site itself. Omitting --wait returns a job ID for later status checks or cancellation; Alexandria settings carry over to follow-ups in the same thread.
What it helps with
- collect pricing tiers from a website
- extract products into a JSON schema
- read feature lists from selected URLs
- check a provider before autonomous extraction
- return a job ID for later polling
Example prompts for the agent
- Extract all pricing tiers from the site and save the result to
.firecrawl/pricing.jsonwith--wait --json -o. - Extract products using a JSON schema with
nameandprice, then write the result to.firecrawl/products.json. - Get the feature list from
<url>only and return structured JSON.
How to install firecrawl-agent
1Skills CLI (Vercel)
npx skills add firecrawl/cli --skill firecrawl-agent -a claude-code2Manually: put the files in place
~/.claude/skills/firecrawl-agent/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/firecrawl-agent/
1Skills CLI (Vercel)
npx skills add firecrawl/cli --skill firecrawl-agent -a codex2Manually: put the files in place
.agents/skills/firecrawl-agent/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add firecrawl/cli --skill firecrawl-agent -a cursor2Manually: put the files in place
.cursor/skills/firecrawl-agent/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install firecrawl/cli firecrawl-agent2Manually: put the files in place
.github/skills/firecrawl-agent/SKILL.md1Skills CLI (Vercel)
npx skills add firecrawl/cli --skill firecrawl-agent -a gemini-cli2Platform CLI
gemini skills install https://github.com/firecrawl/cli --path skills/firecrawl-agentManually: put the files in place
.devin/skills/firecrawl-agent/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add firecrawl/cli --skill firecrawl-agent -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-agent
How do I install firecrawl-agent in Claude Code?
Run npx skills add firecrawl/cli --skill firecrawl-agent -a claude-code, then check ~/.claude/skills/firecrawl-agent/SKILL.md.
What does firecrawl-agent do?
It navigates website pages autonomously and extracts structured data. Results can be returned as JSON or handled later through a job ID.
How is firecrawl-agent different from scrape?
Use scrape for a simple single-page extraction. firecrawl-agent handles autonomous navigation and multi-page collection, while consuming more credits.
