Skill + CLI

Firecrawl Skill is an easy way for AI agents such as Claude Code, Antigravity and OpenCode to use Firecrawl through the CLI.

Installation

If you are using any AI agent like Claude Code, you can install the Firecrawl skill below and the agent will be able to set it up for you.

npx -y firecrawl-cli@latest init --all --browser

• --all installs the Firecrawl skill to every detected AI coding agent
• --browser opens the browser for Firecrawl authentication automatically

After installing the skill, restart your agent for it to discover the new skill.

You can also manually install the Firecrawl CLI globally using npm:

CLI

# Install globally with npm
npm install -g firecrawl-cli

Authentication

Before using the CLI, you need to authenticate with your Firecrawl API key.

Login

CLI

# Interactive login (opens browser or prompts for API key)
firecrawl login

# Login with browser authentication (recommended for agents)
firecrawl login --browser

# Login with API key directly
firecrawl login --api-key fc-YOUR-API-KEY

# Or set via environment variable
export FIRECRAWL_API_KEY=fc-YOUR-API-KEY

View Configuration

CLI

# View current configuration and authentication status
firecrawl view-config

Logout

CLI

# Clear stored credentials
firecrawl logout

Self-Hosted / Local Development

For self-hosted Firecrawl instances or local development, use the --api-url option:

CLI

# Use a local Firecrawl instance (no API key required)
firecrawl --api-url http://localhost:3002 scrape https://example.com

# Or set via environment variable
export FIRECRAWL_API_URL=http://localhost:3002
firecrawl scrape https://example.com

# Configure and persist the custom API URL
firecrawl config --api-url http://localhost:3002

When using a custom API URL (anything other than https://api.firecrawl.dev), API key authentication is automatically skipped, allowing you to use local instances without an API key.

Check Status

Verify installation, authentication, and view rate limits:

CLI

firecrawl --status

Output when ready:

  🔥 firecrawl cli v1.1.1

  ● Authenticated via FIRECRAWL_API_KEY
  Concurrency: 0/100 jobs (parallel scrape limit)
  Credits: 500,000 remaining

• Concurrency: Maximum parallel jobs. Run parallel operations close to this limit but not above.
• Credits: Remaining API credits. Each scrape/crawl consumes credits.

Commands

Scrape

Scrape a single URL and extract its content in various formats.

Use --only-main-content to get clean output without navigation, footers, and ads. This is recommended for most use cases where you want just the article or main page content.

CLI

# Scrape a URL (default: markdown output)
firecrawl https://example.com

# Or use the explicit scrape command
firecrawl scrape https://example.com

# Recommended: use --only-main-content for clean output without nav/footer
firecrawl https://example.com --only-main-content

Output Formats

CLI

# Get HTML output
firecrawl https://example.com --html

# Multiple formats (returns JSON)
firecrawl https://example.com --format markdown,links

# Get images from a page
firecrawl https://example.com --format images

# Get a summary of the page content
firecrawl https://example.com --format summary

# Track changes on a page
firecrawl https://example.com --format changeTracking

# Available formats: markdown, html, rawHtml, links, screenshot, json, images, summary, changeTracking, attributes, branding

Scrape Options

CLI

# Extract only main content (removes navs, footers)
firecrawl https://example.com --only-main-content

# Wait for JavaScript rendering
firecrawl https://example.com --wait-for 3000

# Take a screenshot
firecrawl https://example.com --screenshot

# Include/exclude specific HTML tags
firecrawl https://example.com --include-tags article,main
firecrawl https://example.com --exclude-tags nav,footer

# Save output to file
firecrawl https://example.com -o output.md

# Pretty print JSON output
firecrawl https://example.com --format markdown,links --pretty

# Force JSON output even with single format
firecrawl https://example.com --json

# Show request timing information
firecrawl https://example.com --timing

Available Options:

Option	Short	Description
--url <url>	-u	URL to scrape (alternative to positional argument)
--format <formats>	-f	Output formats (comma-separated): markdown, html, rawHtml, links, screenshot, json, images, summary, changeTracking, attributes, branding
--html	-H	Shortcut for --format html
--only-main-content		Extract only main content
--wait-for <ms>		Wait time in milliseconds for JS rendering
--screenshot		Take a screenshot
--include-tags <tags>		HTML tags to include (comma-separated)
--exclude-tags <tags>		HTML tags to exclude (comma-separated)
--output <path>	-o	Save output to file
--json		Force JSON output even with single format
--pretty		Pretty print JSON output
--timing		Show request timing and other useful information

---

Search

Search the web and optionally scrape the results.

CLI

# Search the web
firecrawl search "web scraping tutorials"

# Limit results
firecrawl search "AI news" --limit 10

# Pretty print results
firecrawl search "machine learning" --pretty

Search Options

CLI

# Search specific sources
firecrawl search "AI" --sources web,news,images

# Search with category filters
firecrawl search "react hooks" --categories github
firecrawl search "machine learning" --categories research,pdf

# Time-based filtering
firecrawl search "tech news" --tbs qdr:h   # Last hour
firecrawl search "tech news" --tbs qdr:d   # Last day
firecrawl search "tech news" --tbs qdr:w   # Last week
firecrawl search "tech news" --tbs qdr:m   # Last month
firecrawl search "tech news" --tbs qdr:y   # Last year

# Location-based search
firecrawl search "restaurants" --location "Berlin,Germany" --country DE

# Search and scrape results
firecrawl search "documentation" --scrape --scrape-formats markdown

# Save to file
firecrawl search "firecrawl" --pretty -o results.json

Available Options:

Option	Description
--limit <number>	Maximum results (default: 5, max: 100)
--sources <sources>	Sources to search: web, images, news (comma-separated)
--categories <categories>	Filter by category: github, research, pdf (comma-separated)
--tbs <value>	Time filter: qdr:h (hour), qdr:d (day), qdr:w (week), qdr:m (month), qdr:y (year)
--location <location>	Geo-targeting (e.g., “Berlin,Germany”)
--country <code>	ISO country code (default: US)
--timeout <ms>	Timeout in milliseconds (default: 60000)
--ignore-invalid-urls	Exclude URLs invalid for other Firecrawl endpoints
--scrape	Scrape search results
--scrape-formats <formats>	Formats for scraped content (default: markdown)
--only-main-content	Include only main content when scraping (default: true)
--json	Output as JSON
--output <path>	Save output to file
--pretty	Pretty print JSON output

---

Map

Discover all URLs on a website quickly.

CLI

# Discover all URLs on a website
firecrawl map https://example.com

# Output as JSON
firecrawl map https://example.com --json

# Limit number of URLs
firecrawl map https://example.com --limit 500

Map Options

CLI

# Filter URLs by search query
firecrawl map https://example.com --search "blog"

# Include subdomains
firecrawl map https://example.com --include-subdomains

# Control sitemap usage
firecrawl map https://example.com --sitemap include   # Use sitemap
firecrawl map https://example.com --sitemap skip      # Skip sitemap
firecrawl map https://example.com --sitemap only      # Only use sitemap

# Ignore query parameters (dedupe URLs)
firecrawl map https://example.com --ignore-query-parameters

# Wait for map to complete with timeout
firecrawl map https://example.com --wait --timeout 60

# Save to file
firecrawl map https://example.com -o urls.txt
firecrawl map https://example.com --json --pretty -o urls.json

Available Options:

Option	Description
--url <url>	URL to map (alternative to positional argument)
--limit <number>	Maximum URLs to discover
--search <query>	Filter URLs by search query
--sitemap <mode>	Sitemap handling: include, skip, only
--include-subdomains	Include subdomains
--ignore-query-parameters	Treat URLs with different params as same
--wait	Wait for map to complete
--timeout <seconds>	Timeout in seconds
--json	Output as JSON
--output <path>	Save output to file
--pretty	Pretty print JSON output

---

Browser

Have your agents interact with the web using a secure browser sandbox.

Launch cloud browser sessions and execute Python, JavaScript, or bash code remotely. Each session runs a full Chromium instance — no local browser install required. Code runs server-side with a pre-configured Playwright page object ready to use.

CLI

# Launch a cloud browser session
firecrawl browser launch-session

# Execute agent-browser commands (default - "agent-browser" is auto-prefixed)
firecrawl browser execute "open https://example.com"
firecrawl browser execute "snapshot"
firecrawl browser execute "click @e5"
firecrawl browser execute "scrape"

# Execute Playwright Python code
firecrawl browser execute --python 'await page.goto("https://example.com")
print(await page.title())'

# Execute Playwright JavaScript code
firecrawl browser execute --node 'await page.goto("https://example.com"); console.log(await page.title());'

# List all sessions (or: list active / list destroyed)
firecrawl browser list

# Close the active session
firecrawl browser close

Browser Options

CLI

# Launch with custom TTL (10 minutes) and live view
firecrawl browser launch-session --ttl 600 --stream

# Launch with inactivity timeout
firecrawl browser launch-session --ttl 120 --ttl-inactivity 60

# agent-browser commands (default - "agent-browser" is auto-prefixed)
firecrawl browser execute "open https://news.ycombinator.com"
firecrawl browser execute "snapshot"
firecrawl browser execute "click @e3"
firecrawl browser execute "scrape"

# Playwright Python - navigate, interact, extract
firecrawl browser execute --python '
await page.goto("https://news.ycombinator.com")
items = await page.query_selector_all(".titleline > a")
for item in items[:5]:
    print(await item.text_content())
'

# Playwright JavaScript - same page object
firecrawl browser execute --node '
await page.goto("https://example.com");
const title = await page.title();
console.log(title);
'

# Explicit bash mode - runs in the sandbox
firecrawl browser execute --bash "agent-browser snapshot"

# Target a specific session
firecrawl browser execute --session <id> --python 'print(await page.title())'

# Save output to file
firecrawl browser execute "scrape" -o result.txt

# Close a specific session
firecrawl browser close --session <id>

# List sessions (all / active / destroyed)
firecrawl browser list
firecrawl browser list active --json

Subcommands:

Subcommand	Description
launch-session	Launch a new cloud browser session (returns session ID, CDP URL, and live view URL)
execute <code>	Execute Playwright Python/JS code or bash commands in a session
list [status]	List browser sessions (filter by active or destroyed)
close	Close a browser session

Execute Options:

Option	Description
--bash	Execute bash commands remotely in the sandbox (default). agent-browser (40+ commands) is pre-installed and auto-prefixed. CDP_URL is auto-injected so agent-browser connects to your session automatically. Best approach for AI agents.
--python	Execute as Playwright Python code. A Playwright page object is available — use await page.goto(), await page.title(), etc.
--node	Execute as Playwright JavaScript code. Same page object available.
--session <id>	Target a specific session (default: active session)

Launch Options:

Option	Description
--ttl <seconds>	Total session TTL (default: 600, range: 30–3600)
--ttl-inactivity <seconds>	Auto-close after inactivity (range: 10–3600)
--profile <name>	Name for a profile (saves and reuses browser state across sessions)
--no-save-changes	Load existing profile data without saving changes back
--stream	Enable live view streaming

Common Options:

Option	Description
--output <path>	Save output to file
--json	Output as JSON format

---

Crawl

Crawl an entire website starting from a URL.

CLI

# Start a crawl (returns job ID immediately)
firecrawl crawl https://example.com

# Wait for crawl to complete
firecrawl crawl https://example.com --wait

# Wait with progress indicator
firecrawl crawl https://example.com --wait --progress

Check Crawl Status

CLI

# Check crawl status using job ID
firecrawl crawl <job-id>

# Example with a real job ID
firecrawl crawl 550e8400-e29b-41d4-a716-446655440000

Crawl Options

CLI

# Limit crawl depth and pages
firecrawl crawl https://example.com --limit 100 --max-depth 3 --wait

# Include only specific paths
firecrawl crawl https://example.com --include-paths /blog,/docs --wait

# Exclude specific paths
firecrawl crawl https://example.com --exclude-paths /admin,/login --wait

# Include subdomains
firecrawl crawl https://example.com --allow-subdomains --wait

# Crawl entire domain
firecrawl crawl https://example.com --crawl-entire-domain --wait

# Rate limiting
firecrawl crawl https://example.com --delay 1000 --max-concurrency 2 --wait

# Custom polling interval and timeout
firecrawl crawl https://example.com --wait --poll-interval 10 --timeout 300

# Save results to file
firecrawl crawl https://example.com --wait --pretty -o results.json

Available Options:

Option	Description
--url <url>	URL to crawl (alternative to positional argument)
--wait	Wait for crawl to complete
--progress	Show progress indicator while waiting
--poll-interval <seconds>	Polling interval (default: 5)
--timeout <seconds>	Timeout when waiting
--status	Check status of existing crawl job
--limit <number>	Maximum pages to crawl
--max-depth <number>	Maximum crawl depth
--include-paths <paths>	Paths to include (comma-separated)
--exclude-paths <paths>	Paths to exclude (comma-separated)
--sitemap <mode>	Sitemap handling: include, skip, only
--allow-subdomains	Include subdomains
--allow-external-links	Follow external links
--crawl-entire-domain	Crawl entire domain
--ignore-query-parameters	Treat URLs with different params as same
--delay <ms>	Delay between requests
--max-concurrency <n>	Maximum concurrent requests
--output <path>	Save output to file
--pretty	Pretty print JSON output

---

Agent

Search and gather data from the web using natural language prompts.

CLI

# Basic usage - URLs are optional
firecrawl agent "Find the top 5 AI startups and their funding amounts" --wait

# Focus on specific URLs
firecrawl agent "Compare pricing plans" --urls https://slack.com/pricing,https://teams.microsoft.com/pricing --wait

# Use a schema for structured output
firecrawl agent "Get company information" --urls https://example.com --schema '{"name": "string", "founded": "number"}' --wait

# Use schema from a file
firecrawl agent "Get product details" --urls https://example.com --schema-file schema.json --wait

Agent Options

CLI

# Use Spark 1 Pro for higher accuracy
firecrawl agent "Competitive analysis across multiple domains" --model spark-1-pro --wait

# Set max credits to limit costs
firecrawl agent "Gather contact information from company websites" --max-credits 100 --wait

# Check status of an existing job
firecrawl agent <job-id> --status

# Custom polling interval and timeout
firecrawl agent "Summarize recent blog posts" --wait --poll-interval 10 --timeout 300

# Save output to file
firecrawl agent "Find pricing information" --urls https://example.com --wait -o pricing.json --pretty

Available Options:

Option	Description
--urls <urls>	Optional list of URLs to focus the agent on (comma-separated)
--model <model>	Model to use: spark-1-pro (default, higher accuracy) or spark-1-mini (60% cheaper)
--schema <json>	JSON schema for structured output (inline JSON string)
--schema-file <path>	Path to JSON schema file for structured output
--max-credits <number>	Maximum credits to spend (job fails if limit reached)
--status	Check status of existing agent job
--wait	Wait for agent to complete before returning results
--poll-interval <seconds>	Polling interval when waiting (default: 5)
--timeout <seconds>	Timeout when waiting (default: no timeout)
--output <path>	Save output to file
--json	Output as JSON format

---

Credit Usage

Check your team’s credit balance and usage.

CLI

# View credit usage
firecrawl credit-usage

# Output as JSON
firecrawl credit-usage --json --pretty

---

Version

Display the CLI version.

CLI

firecrawl version
# or
firecrawl --version

Global Options

These options are available for all commands:

Option	Short	Description
--status		Show version, auth, concurrency, and credits
--api-key <key>	-k	Override stored API key for this command
--api-url <url>		Use custom API URL (for self-hosted/local development)
--help	-h	Show help for a command
--version	-V	Show CLI version

Output Handling

The CLI outputs to stdout by default, making it easy to pipe or redirect:

CLI

# Pipe markdown to another command
firecrawl https://example.com | head -50

# Redirect to a file
firecrawl https://example.com > output.md

# Save JSON with pretty formatting
firecrawl https://example.com --format markdown,links --pretty -o data.json

Format Behavior

• Single format: Outputs raw content (markdown text, HTML, etc.)
• Multiple formats: Outputs JSON with all requested data

CLI

# Raw markdown output
firecrawl https://example.com --format markdown

# JSON output with multiple formats
firecrawl https://example.com --format markdown,links

Examples

Quick Scrape

CLI

# Get markdown content from a URL (use --only-main-content for clean output)
firecrawl https://docs.firecrawl.dev --only-main-content

# Get HTML content
firecrawl https://example.com --html -o page.html

Full Site Crawl

CLI

# Crawl a docs site with limits
firecrawl crawl https://docs.example.com --limit 50 --max-depth 2 --wait --progress -o docs.json

Site Discovery

CLI

# Find all blog posts
firecrawl map https://example.com --search "blog" -o blog-urls.txt

Research Workflow

CLI

# Search and scrape results for research
firecrawl search "machine learning best practices 2024" --scrape --scrape-formats markdown --pretty

Agent

CLI

# URLs are optional
firecrawl agent "Find the top 5 AI startups and their funding amounts" --wait

# Focus on specific URLs
firecrawl agent "Compare pricing plans" --urls https://slack.com/pricing,https://teams.microsoft.com/pricing --wait

Browser Automation

CLI

# Launch a session, scrape a page, and close
firecrawl browser launch-session
firecrawl browser execute "open https://news.ycombinator.com"
firecrawl browser execute "snapshot"
firecrawl browser execute "scrape"
firecrawl browser close

# Use agent-browser via bash mode (default — recommended for AI agents)
firecrawl browser launch-session
firecrawl browser execute "open https://example.com"
firecrawl browser execute "snapshot"
# snapshot returns @ref IDs — use them to interact
firecrawl browser execute "click @e5"
firecrawl browser execute "fill @e3 'search query'"
firecrawl browser execute "scrape"
# Run --help to see all 40+ commands
firecrawl browser execute --bash "agent-browser --help"
firecrawl browser close

Combine with Other Tools

CLI

# Extract URLs from search results
jq -r '.data.web[].url' search-results.json

# Get titles from search results
jq -r '.data.web[] | "\(.title): \(.url)"' search-results.json

# Extract links and process with jq
firecrawl https://example.com --format links | jq '.links[].url'

# Count URLs from map
firecrawl map https://example.com | wc -l

Telemetry

The CLI collects anonymous usage data during authentication to help improve the product:

• CLI version, OS, and Node.js version
• Development tool detection (e.g., Cursor, VS Code, Claude Code)

No command data, URLs, or file contents are collected via the CLI.

To disable telemetry, set the environment variable:

CLI

export FIRECRAWL_NO_TELEMETRY=1

Open Source

The Firecrawl CLI and Skill are open source and available on GitHub: firecrawl/cli