Skip to content

Firecrawl Skills - Setup Guide

Source: firecrawl - official publisher (4 skill repos) GitHub: firecrawl/cli · firecrawl/skills · firecrawl/firecrawl-workflows · firecrawl/anydoc Skills: 72 skills across 4 repos (~1.47M combined installs) Category: Web Scraping / Research / Growth Operations First Seen: August 14, 2026 afternoon sweep Quality Tier: 🟢 Production (official vendor skills, 5 skills above 57K installs)

Firecrawl - the web-scraping engine that already powers Hermes's web_search and web_extract backends - ships its own official skill ecosystem on skills.sh. Four repos turn raw Firecrawl API access into repeatable agent workflows: firecrawl/cli for programmatic scrape/search/crawl operations, firecrawl/skills for building scraping integrations, firecrawl-workflows for end-to-end research pipelines (deep research, market research, SEO audit, lead gen, competitive intel), and firecrawl/anydoc for converting documents to markdown.

Installation

# Core CLI operations (scrape, search, crawl, map, parse, monitor, browser, interact)
npx skills add firecrawl/cli

# Build skills (scraping integration development + onboarding)
npx skills add firecrawl/skills

# Research workflows (deep research, market research, SEO audit, lead gen, competitive intel)
npx skills add firecrawl/firecrawl-workflows

# Any-to-markdown document conversion
npx skills add firecrawl/anydoc

Set the API key in the environment:

export FIRECRAWL_API_KEY="fc-..."

A free tier is available (trial credits); paid tiers add higher rate limits, concurrency, and advanced extraction (JSON schema output, interact, monitor).

Prerequisites

Requirement Details
Firecrawl API key FIRECRAWL_API_KEY env var - free tier available at firecrawl.dev
Hermes Agent Any recent version; the skills are plain markdown workflows + CLI commands
npx / Node.js For installing skills via npx skills add
Hermes web tools (optional) Hermes's own web_search / web_extract are Firecrawl-backed via the Nous subscription - these skills add direct API control on top

What It Provides

firecrawl/cli (14 skills, 716.0K installs)

Skill Installs Purpose
firecrawl 97.8K Core Firecrawl operations: scrape, search, crawl, map, extract
firecrawl-search 79.1K Web search with Firecrawl's search API
firecrawl-scrape 78.9K Single-URL scraping with markdown/JSON output
firecrawl-crawl 76.9K Multi-page crawls with depth and path filters
firecrawl-agent 76.8K Agent-oriented usage patterns for the Firecrawl API
firecrawl-map 76.5K Site URL discovery (sitemap-style mapping)
firecrawl-download 76.2K Download page content/assets
firecrawl-interact 61.6K Browser-interaction scraping (clicks, forms, JS rendering)
firecrawl-parse 47.7K Structured extraction via JSON schema
firecrawl-monitor 30.0K Recurring page monitoring for changes
firecrawl-browser 11.4K Full browser session control for hard-to-scrape pages
firecrawl-instruct 3.3K Natural-language extraction instructions

firecrawl-workflows (16 skills, 492.8K installs)

End-to-end pipelines, each 30K+ installs: firecrawl-deep-research (32.6K), firecrawl-research-papers (31.8K), firecrawl-website-design-clone (31.5K), firecrawl-market-research (31.1K), firecrawl-seo-audit (30.8K), firecrawl-knowledge-base (30.7K), firecrawl-dashboard-reporting (30.6K), firecrawl-workflows (30.5K), firecrawl-lead-gen (30.5K), firecrawl-lead-research (30.5K), firecrawl-shop (30.4K), firecrawl-qa (30.4K), firecrawl-competitive-intel (30.4K), firecrawl-knowledge-ingest (30.4K), firecrawl-company-directories (30.4K), firecrawl-demo-walkthrough (30.4K).

firecrawl/skills (41 skills, 252.5K installs)

Build-time skills for integrating Firecrawl into agent apps: firecrawl-build-scrape (58.1K), firecrawl-build-search (58.0K), firecrawl-build-interact (57.8K), firecrawl-build-onboarding (57.7K), firecrawl-research-index (18.1K), plus deploy/QA/tooling skills (vercel/netlify/render/cloudflare deploys, Notion integration, GitHub CI fixes, playwright, pdf, spreadsheet, screenshot).

firecrawl/anydoc (1 skill, ~4K installs)

convert-documents-to-markdown - convert any document (PDF, docx, xlsx, images) to clean markdown for agent ingestion.

Quick Start

  1. export FIRECRAWL_API_KEY="fc-..." and npx skills add firecrawl/cli
  2. "Scrape https://example.com/pricing and return the markdown"
  3. "Search the web for 'AI agent platforms' and summarize the top 10 results"
  4. "Run a deep research pass on multi-agent orchestration tooling, cite every claim"
  5. "Map every URL on our competitor's site and flag pages mentioning pricing"

CorpusIQ Use Cases

Use Case How
Competitive intel firecrawl-competitive-intel + firecrawl-seo-audit against competitor domains before feature/positioning decisions
Market research firecrawl-market-research for operator pain-point mining - feeds the help-first content engine
Lead research firecrawl-lead-research + firecrawl-company-directories to enrich cold-outreach targets with real site data
Docs SEO firecrawl-seo-audit as a second opinion alongside the Search Console/Ahrefs/Semrush connector stack
Knowledge ingest firecrawl-anydoc to convert partner PDFs/research into markdown for the GBrain vault

Integration with CorpusIQ

CorpusIQ is the connector layer that lets AI assistants read the systems a business already runs on - accounting, payments, CRM, email, files and analytics - and answer with the source named. It is read-only by design: it retrieves data and never writes to, or acts in, a connected system, which is what makes it safe to point at live accounts.

This page is part of the CorpusIQ directory, the reference we maintain for the MCP server and agent-skill ecosystem. Run alongside CorpusIQ, a third-party server like this one can be combined with a company's own business data in a single answer.

See what CorpusIQ connects

Limitations / Verification

  • API-key based: usage is metered against your Firecrawl plan; monitor credit consumption on long crawls
  • interact and browser consume more credits than plain scrape/search
  • monitor requires a paid plan for recurring schedules
  • Skills are agent-agnostic markdown - they work in Hermes sessions the same way as Claude Code
# Verify key works (free tier)
curl -s -H "Authorization: Bearer $FIRECRAWL_API_KEY" https://api.firecrawl.dev/v1/scrape \
  -d '{"url":"https://example.com","formats":["markdown"]}' -H "Content-Type: application/json" | head -c 300

← Skills Catalog | Marketplace →

Powered by CorpusIQ

AI answers on your data. See where every number comes from.

Connect your tools and ask in ChatGPT, Claude, or Perplexity. Enterprise-grade answers in minutes. From $29.95/month.

Start free, no credit card
Contact Terms Privacy Pricing