firecrawl Skill: Let Your Assistant Search, Scrape and Crawl the Web

Makes your coding assistant use the Firecrawl command line tool for web search, page reading, crawling and change alerts.

Installs: 103.8K; on skills.sh
Stars: 676; on GitHub
Files: 2; besides SKILL.md
Updated: Oct 2026; last repo commit

The firecrawl skill is the main entry for the Firecrawl command line tool. It tells your assistant which command to use for a web task: search, scrape, map, crawl, agent, interact, parse, download or monitor. It also tells the assistant to use Firecrawl for ordinary web research, even when you do not name it.

The skill only gives instructions. You must install the Firecrawl CLI and sign in. Results are saved as files in a .firecrawl/ folder, and each call can spend Firecrawl credits.

Good for

Developers who want Claude Code or Codex to read live web pages and docs
Researchers who gather sources and save them as local files
Teams that watch pricing, jobs or docs pages for changes
Anyone who needs structured data pulled from a messy site

Install firecrawl

Claude Code

Copy this prompt.

Install the firecrawl skill in Claude Code. Run this command:

npm install -g firecrawl-cli@latest
firecrawl login --browser
npx skills add https://github.com/firecrawl/cli --skill '*' -a claude-code

Then check that it loaded and tell me how to use it.

2
Paste it into Claude Code. It shows the change and asks before running it.

Requires Node.js, A Firecrawl account (or the keyless free tier).

The skill only gives your assistant instructions. The Firecrawl command line tool does the work, so install it and sign in first. Search, scrape and interact also run without an account on a rate-limited free tier, according to the install rules. Crawl, map, download, agent and monitor need an account. The skills command above installs every skill in the Firecrawl CLI repo, which the main skill routes between. The one-command route installs the CLI, signs you in through the browser and adds skills for every editor it detects, not just this skill. The install rules say it adds core and workflow skills. The README says the CLI skills are the default and the workflow skills are an optional multi-select in interactive setup, so a non-interactive -y run may not include them.

From the vendor's docs. Source.

What's in the skill

When it loads

Any live-web task via the Firecrawl CLI — including ordinary web research: searching the web, reading or extracting pages, gathering sources, discovering site URLs, bulk extraction, downloading a site, change alerts, or pages needing clicks/login — web only; local files route to firecrawl-parse. For US legal or regulatory questions use gov; for papers use firecrawl-research-index; for library, API, error, or bug questions use firecrawl-developer-index.

Sections in SKILL.md

Prerequisites
Workflow
Large results and Alexandria
When to Load References
Output & Organization
Feedback
Parallelization
Credit Usage

Key facts

Author: firecrawl org
Repo: firecrawl/cli
Installs on skills.sh: 103,839 as of 2026-11-02
GitHub stars: 676 for the whole firecrawl/cli repo, 2026-11-02
License: ISC (package.json of the CLI repo)
Last commit to the repo: 2026-10-28
What is inside: 3 files including SKILL.md (rules/install.md and rules/security.md)
Needs: The Firecrawl CLI (npm package firecrawl-cli) and Node.js. Search, scrape and interact work without an account on a rate-limited free tier; crawl, map, download, agent and monitor need an account.
Related skills: firecrawl-search, firecrawl-scrape, firecrawl-map, firecrawl-crawl, firecrawl-agent, firecrawl-interact, firecrawl-download, firecrawl-parse, firecrawl-monitor, plus the research, developer and Alexandria skills in the same repo
skills.sh audits: Gen Agent Trust Hub pass, Socket pass, Snyk warn (2026-11-02)

What it includes

Search, scrape, map, crawl, monitor, then interact
The skill tells the assistant to start with the real question, read known URLs directly, find a subpage with map, take a whole section with crawl, set a monitor for recurring checks, and only then drive the page with interact.
Try: “Find the pricing page for example.com and tell me what the enterprise plan costs.”
Runs code · Escalation pattern
Finds pages on a topic when you have no URL
Detail lives in the firecrawl-search skill. The main skill also lists research, developer and gov commands for papers, coding questions and US law.
Runs code · search
Reads one page or a list of known URLs
Works for static and JavaScript pages. batch-scrape is for many URLs, extra formats, JSON extraction or a webhook.
Runs code · scrape and batch-scrape
Lists the URLs on a site, with search filtering
Use it to find a specific subpage before scraping it.
Runs code · map
Pulls many pages from one site section
For jobs such as everything under /docs.
Runs code · crawl
AI extraction of structured data from complex sites
For data that needs navigation across pages.
Runs code · agent
Clicks, form fills, pagination and login on a page
Scrape first, then drive the live page.
Runs code · scrape + interact
Saves a site as local files
Experimental command under firecrawl x.
Runs code · x download
Turns a local PDF, DOCX or XLSX into markdown
For a file path rather than a URL.
Runs code · parse
Alerts you by email or webhook when a page changes
For recurring checks instead of repeated one-off scrapes.
Runs code · monitor
Writes results to .firecrawl/ and reads them in small pieces
Results go to files with -o, the folder is added to .gitignore, and the assistant reads them with head or grep. URLs are quoted in shell commands.
Write · Output rules
Shows sign-in state, concurrency and remaining credits
firecrawl --status is the first check the skill runs. credit-usage needs an account.
Runs code · credit-usage and --status

From the official docs

Permissions

Medium risk

What it can do

Read your data
Create and change things
Delete things
Read-only mode

Watch out for

It spends Firecrawl credits
Search, scrape, crawl, agent and monitor calls use credits. Check firecrawl credit-usage before large crawls and ask the assistant to cap the size of a job. github.com
It defaults to Firecrawl for web research
The skill tells the assistant to use Firecrawl for ordinary web research even when you do not name it. Remove the skill if you want another search route. github.com
It writes files and edits .gitignore
Results are saved under .firecrawl/ and the skill says to add that folder to .gitignore. Check the diff before you commit. github.com
It can send feedback to Firecrawl
The skill describes feedback commands for searches and jobs, with ratings, issue codes and short notes. Set FIRECRAWL_NO_ENDPOINT_FEEDBACK=1 to skip them. github.com
Web pages can try to steer your assistant
The skill's own security note calls fetched web content untrusted and says it may contain prompt injection. Do not let the assistant follow instructions it finds on a page. github.com
Setup installs a global npm package
The CLI is installed with npm install -g. The one-command setup also puts skills in the shared ~/.agents/skills folder and makes them available to every coding editor it detects. The README says it always includes Codex as a target, even when Codex is not detected. Use --agent <name> to limit it, or install skills by name with the skills command. github.com
The CLI sends anonymous usage data
The README says it collects the CLI version, OS, Node.js version and which dev tools it detects during sign-in. Set FIRECRAWL_NO_TELEMETRY=1 to turn it off. github.com

Security advisories

No published security advisories as of Nov 2, 2026.

Technical details

Turning on read-only	No read-only mode. The skill runs CLI commands that spend credits and write files.
Talks to	Firecrawl's API, or your own server if you set --api-url. Web pages the CLI fetches. Page content the assistant reads goes to your AI provider as part of the chat
Sign-in and keys	Browser sign-in stores credentials in the CLI. You can set FIRECRAWL_API_KEY instead.
Scopes	Runs firecrawl commands in your shell (the skill allows Bash(firecrawl ) and Bash(npx firecrawl-cli ))

Questions

What does the firecrawl skill do?

It tells your assistant how to use the Firecrawl CLI for web work. It picks between search, scrape, map, crawl, agent, interact, parse, download and monitor, and saves results as files in .firecrawl/.

How do I install the firecrawl skill?

Use the install box at the top of this page. You need the Firecrawl CLI and a sign-in as well as the skill. The first route installs all skills in the repo, because this one hands off to the others.

Do I need a Firecrawl account or API key?

Not for search, scrape and interact. The install rules say they work on a rate-limited free tier without a key. Crawl, map, download, agent and monitor need an account. Browser login or an API key gives the best results.

Is the firecrawl skill safe to use?

It runs Firecrawl commands in your shell, writes files to .firecrawl/ and spends credits, so watch large crawls. Fetched pages are untrusted text and can carry prompt injection. Do not let the assistant follow instructions it finds on a page.

Why does my assistant use Firecrawl without being asked?

The skill tells it to use Firecrawl for ordinary web research, such as searching the web or reading pages, even when you do not name Firecrawl. Remove the skill if you do not want that.

Does it work in Codex and the Claude app?

The README says the one-command setup always includes Codex as a target. It does not document uploading the skill to the Claude app. The skill runs CLI commands, so it fits an assistant that can run commands on your machine.

Skill or Firecrawl MCP server?

They are two routes to Firecrawl. The firecrawl skill is a SKILL.md that tells the assistant to run the Firecrawl CLI in your shell. The skill only allows Bash(firecrawl *) and Bash(npx firecrawl-cli *), so it needs an assistant that can run shell commands. The Firecrawl MCP server and the local Firecrawl MCP server give your client Firecrawl as tools it calls directly, with no skill involved. The README describes firecrawl setup mcp for adding the MCP server to your editors. You can use either route, or both.

Can it opt out of feedback calls?

Yes. The skill describes feedback commands that report job results to Firecrawl. Setting FIRECRAWL_NO_ENDPOINT_FEEDBACK=1 makes the CLI skip them, and the skill tells the assistant to respect that flag.

Information

updated Nov 2, 2026

Repository: firecrawl/cli
Folder: skills/firecrawl

firecrawl/cli skills.sh

Sources

Docs, issue trackers and discussions this page is built from.

Show all 5

Firecrawl CLI install rules · github.com
firecrawl/cli README · github.com
skills CLI commands · github.com
firecrawl SKILL.md · github.com
Firecrawl security rules · github.com