Skip to main content
skill Claude Code CX Codex Cursor GitHub Copilot

firecrawl-workflows — Claude Code skill: what it does, install

A router skill for Claude Code and other coding agents: it picks a ready Firecrawl workflow for the job and returns a sourced deliverable.

Install into Claude Code · Skills CLI (Vercel) All methods and platforms ↓
npx skills add firecrawl/firecrawl-workflows --skill firecrawl-workflows -a claude-code
Open on GitHub

What firecrawl-workflows does

Think of firecrawl-workflows as a dispatcher. Hand the agent a job, and it routes to the matching Firecrawl workflow: lead discovery and qualification, lead CSV enrichment, business contacts, current job postings, SEO audits, QA on a live site, brand and UI extraction, docs turned into a knowledge base. Evidence gets collected through the Firecrawl CLI or API, and claims keep their source links.

Every branch carries its own deliverable and its own stopping point. A contact card ends as a contact card, with no research report bolted on. With no matching branch, the skill falls back to a generic loop: confirm the goal, gather web evidence, split independent units across sub-agents, then synthesize. It closes with a short rerun-inputs block for repeatable jobs.

CLI and API runs need FIRECRAWL_API_KEY. No key, and the workflow dies on the first request, which is the point to check your .env before briefing the agent. Notes on writing new branches sit in references/workflow-authoring.md.

What it helps with

  • extract a filtered company list from a directory
  • run QA on a live site and file a bug report
  • verify which job postings are still open
  • build a meeting brief on a supplied company
  • chunk a docs site into RAG-ready files

Example prompts for the agent

  • Audit example.com for SEO: site structure, meta tags, SERP. Give me a report with source links.
  • Walk the checkout flow on the test site and write up a bug report with repro steps.
  • Pull docs.example.com into a knowledge base formatted for RAG.

How to install firecrawl-workflows

1Skills CLI (Vercel)

npx skills add firecrawl/firecrawl-workflows --skill firecrawl-workflows -a claude-code

2Manually: put the files in place

~/.claude/skills/firecrawl-workflows/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/firecrawl-workflows/

From the same set6

The whole set →

Similar

browser-trace browserbase Attaches a second CDP client to a browser session that is already running, records the whole DevTools stream to disk, and splits it so you can search a failed run. Claude Code CX Codex Cursor +1 3,000installs fetch browserbase Retrieves HTML or JSON from static web pages via Browserbase Fetch API without starting a browser session. Useful for fast scraping and header inspection. Claude Code CX Codex Cursor +1 3,000installs browser-mcp-agent antibrow Gives an agent a real browser through MCP calls: navigate, click, fill forms, screenshot, run JS. A named profile keeps the session logged in between runs. Claude Code CX Codex Cursor +1 105,700installs agent-browser vercel-labs A CLI skill for browser control: it opens pages, fills forms, clicks buttons, and tests web apps. Use it when an agent needs to act inside a website. Claude Code CX Codex Cursor +1 1,200,000installs webapp-testing official anthropics This skill tests local web apps with Playwright, helping agents debug UI behavior, capture screenshots, and inspect browser logs. Claude Code Claude.ai 186,758installs computer-use stablyai This skill drives a visible app window through orca computer: accessibility tree, clicks, typing, screenshots. Reach for it when the job sits behind a GUI. Claude Code CX Codex Cursor +1 239,000installs browser-use browser-use Hands the agent a real Chrome over CDP: clicks, typing, screenshots, and pages that need JS or your logged-in session. Claude Code CX Codex Cursor +1 98,100installs last30days mvanhorn Pulls posts and engagement on any topic from the last 30 days across Reddit, X, YouTube, TikTok, Hacker News, Polymarket, GitHub and the web. Useful when you want what people actually said, not what a blog summarized. Claude Code CX Codex Cursor +1 52,400installs

Questions about firecrawl-workflows

How do I install firecrawl-workflows in Claude Code?

Run npx skills add firecrawl/firecrawl-workflows --skill firecrawl-workflows -a claude-code. The skill file lands at ~/.claude/skills/firecrawl-workflows/SKILL.md.

What does firecrawl-workflows do?

It selects the right Firecrawl workflow for your request and runs it through to an artifact: a report, a lead list, a bug report, a knowledge base. Raw extraction and product-code integration are handled elsewhere.

Does it need a Firecrawl API key?

Yes, FIRECRAWL_API_KEY is required when the workflow runs through the CLI or the API.

Source text

SKILL.md · ISC
---
name: firecrawl-workflows
description: Run focused Firecrawl workflows for research, documentation lookup, SEO audits, QA, lead discovery and enrichment, business contacts, current jobs, fantasy decisions, brand assets, knowledge bases and website design systems. Use when the user wants a sourced answer or deliverable from a workflow rather than raw extraction or product-code integration.
license: ISC
metadata:
  author: firecrawl
  version: "0.1.1"
  homepage: https://www.firecrawl.dev
  source: https://github.com/firecrawl/firecrawl-workflows
inputs:
  - name: FIRECRAWL_API_KEY
    description: Firecrawl API key for hosted Firecrawl requests when the workflow runs through the CLI or API.
    required: true
references:
  - references/workflow-authoring.md
---

# Firecrawl Workflows

Use this when the user wants a finished deliverable powered by Firecrawl, not only raw web extraction and not product-code integration.

## Choose The Workflow

- Use [firecrawl-lead-gen](../firecrawl-lead-gen/SKILL.md) to discover and qualify new professional or business prospects, or extract lists from a specified database or directory.
- Use [firecrawl-lead-research](../firecrawl-lead-research/SKILL.md) to research supplied people or companies, enrich a lead CSV while preserving rows, or prepare a meeting brief.
- Use [firecrawl-business-contacts](../firecrawl-business-contacts/SKILL.md) to find published organizational contacts for named or described businesses.
- Use [firecrawl-job-postings](../firecrawl-job-postings/SKILL.md) to find and verify current job openings.
- Use [firecrawl-fantasy-sports](../firecrawl-fantasy-sports/SKILL.md) for a fantasy decision grounded in current news and league context.
- Use the core `firecrawl-developer-index` skill (installed with the Firecrawl CLI skills) for official documentation, setup, usage, migrations, compatibility and upstream implementation or bug evidence.
- Use [firecrawl-website-design-clone](../firecrawl-website-design-clone/SKILL.md) for an existing website's brand assets or its UI system. Return a compact asset reference for a document, or an agent-ready `DESIGN.md` for UI implementation.
- Use [firecrawl-research-papers](../firecrawl-research-papers/SKILL.md) for literature reviews and paper-backed synthesis, including biomedical, clinical, drug, gene, disease, epidemiology, and public-health topics. It queries Firecrawl's paper index — PubMed, bioRxiv, medRxiv, and arXiv abstracts with full text reachable per paper — rather than searching websites.
- Use [firecrawl-deep-research](../firecrawl-deep-research/SKILL.md) for sourced multi-source research reports built from **web** evidence: market, policy, technical, or industry topics. Not for literature reviews — when the evidence base is published papers, use `firecrawl-research-papers` above.
- Use [firecrawl-seo-audit](../firecrawl-seo-audit/SKILL.md) for site structure, on-page SEO, keyword, and SERP audits.
- Use [firecrawl-qa](../firecrawl-qa/SKILL.md) for live-site QA testing and bug reports.
- Use [firecrawl-competitive-intel](../firecrawl-competitive-intel/SKILL.md) for recurring pricing, feature, and changelog monitoring.
- Use [firecrawl-company-directories](../firecrawl-company-directories/SKILL.md) for directory extraction into company lists.
- Use [firecrawl-dashboard-reporting](../firecrawl-dashboard-reporting/SKILL.md) for dashboard metrics extraction.
- Use [firecrawl-knowledge-base](../firecrawl-knowledge-base/SKILL.md) for LLM-ready docs, RAG chunks, training data, or docs mirrors.
- Use [firecrawl-knowledge-ingest](../firecrawl-knowledge-ingest/SKILL.md) for auth-gated or JS-heavy docs portal ingestion.
- Use [firecrawl-market-research](../firecrawl-market-research/SKILL.md) for market, financial, and industry research.
- Use [firecrawl-demo-walkthrough](../firecrawl-demo-walkthrough/SKILL.md) for product flow walkthroughs and UX teardown reports.
- Use [firecrawl-shop](../firecrawl-shop/SKILL.md) for product research and shopping recommendations.

If no existing workflow fits, use this generic process and produce a reusable pattern that could become a new skill.

For a focused workflow, follow its deliverable and stopping conditions. A contact card, documentation answer or fantasy recommendation does not require a research report. Reuse adequate evidence and preserve explicit source and no-browse instructions.

## Required Intake

Infer the workflow, inputs, audience, and output format from the user's request and surrounding context. If enough is clear, start immediately.

Ask at most 1-3 concise clarifying questions only when a missing input would block the work, such as:

- the URL, company, topic, or source to analyze
- the desired deliverable or output format
- a constraint that would materially change the workflow

Use the host agent's normal way to ask clarifying questions. Do not depend on a harness-specific function name.

## Default Process

1. Identify the matching workflow and requested artifact.
2. Collect web evidence with Firecrawl through the CLI or equivalent Firecrawl tool surface.
3. Save or cite source evidence so the final claims are traceable.
4. Run independent research units in parallel when available.
5. Synthesize findings into the requested deliverable.
6. Include a short "rerun inputs" block when the workflow could be automated.

## Parallel Work

If appropriate, use sub-agents or equivalent parallel task runners for independent units such as:

- one competitor per researcher
- one URL or page per researcher
- one source category per researcher
- one analysis dimension per reviewer

Keep the handoff generic: provide the unit of work, source URLs or search terms, expected extracted fields, and output format.

## Deliverable Standards

Every workflow should return:

- a concise executive summary
- the evidence base used
- the analysis or artifact requested by the user
- recommendations or next actions when useful
- automation inputs for reruns

For authoring new workflow skills, see [workflow-authoring.md](references/workflow-authoring.md).

© firecrawl · ISC · firecrawl/firecrawl-workflows

See also

Project sponsors

Companies that keep this analytics open
Gold sponsor

Impressions, clicks and code copies — open statistics. We keep the numbers public for the same reason NeBlask itself is public. Become a sponsor