Firecrawl Build Skills
Five skills that teach an agent to wire Firecrawl search, scraping and page interaction into the code of your own application
- 201,311 ★ on the repository
- Last commit
- AGPL-3.0
Extended access
Why this level
- The application's requests and page text go to the Firecrawl cloud under your API key.
- The `/interact` skill describes automated actions on pages, including forms and sign-in flows.
Install
- Works in
Manual install
npx skills add firecrawl/skillsThe Firecrawl skills catalog. The build skills from this repository are copied into it automatically.
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- The application's requests and page text go to the Firecrawl cloud under your API key.
- The `/interact` skill describes automated actions on pages, including forms and sign-in flows.
This is third-party code. Review the repository files before installing.
What it does
Firecrawl Build Skills are five skills from the `skills/` folder of the Firecrawl monorepo. They guide an agent when an application needs web data inside its code: choose the right endpoint, install the SDK, get a key and write the call. The root `firecrawl-build` skill analyses the task and picks between `/scrape`, `/search` and `/interact`. Separate skills cover single-page extraction (`firecrawl-build-scrape`), query-based search (`firecrawl-build-search`), clicks, forms and pagination (`firecrawl-build-interact`) and first-time key setup (`firecrawl-build-onboarding`). The skills need a `FIRECRAWL_API_KEY`. For a self-hosted Firecrawl deployment the address goes into `FIRECRAWL_API_URL`.
Who it is for. Developers whose product or agent has to pull data from the web: search, page content, structured fields. The agent writes the integration and you review the code.
Good fit when
- An application gets a feature that turns a link into page text as Markdown or JSON.
- A product starts from a query rather than a URL: web search first, then reading the pages it finds.
- The page returns data only after a click, a form fill or paging through results.
- A project connects Firecrawl for the first time and the agent has to pick an SDK, get a key and store it in `.env`.
- A team runs Firecrawl on its own servers and wants the agent to write code against its own API address.
Not a fit when
- Web data is needed only inside the current agent session: use the Firecrawl CLI, these skills are about application code.
- You want finished reports and audits without writing code: take Firecrawl Workflows.
- There is no Firecrawl account and requests cannot be paid for: the cloud API does not answer without a key.
Example request
Add a function to our service that fetches an article as Markdown by URL through Firecrawl. Pick the endpoint, install the SDK and store the key in `.env`.Limitations
The skills describe integration into code and do not fetch pages themselves. Cloud mode needs an account and a `FIRECRAWL_API_KEY`, and the application's requests go to the Firecrawl service. Firecrawl itself is licensed under AGPL-3.0, which matters for self-hosting. The `firecrawl/skills` catalog is built from this folder automatically and is closed to direct edits. Cloud requests are paid under the Firecrawl plan, so check limits and prices in your account. If a project already uses another scraping tool, the skill will not handle the migration and only suggests call patterns.
How to disable. Delete the `firecrawl-build*` folders from the agent's skills directory. Remove `FIRECRAWL_API_KEY` from the project's `.env` if the integration is no longer needed.
FAQ
How do I install Firecrawl Build Skills?
Run `npx skills add firecrawl/skills`. This is the Firecrawl catalog, into which CI copies the skills from the `skills/` folder of the monorepo.
How do they differ from the Firecrawl CLI?
The CLI gives an agent web data right in its session. Build Skills teach how to embed Firecrawl into application code: endpoint choice, SDK, keys, call patterns.
Do I need a Firecrawl key?
Yes, `FIRECRAWL_API_KEY` is declared required. For self-hosting, also set `FIRECRAWL_API_URL`.
Which endpoint should I choose?
Use `/scrape` when the page address is known, `/search` when you start from a query and `/interact` when clicks and forms are needed. The root `firecrawl-build` skill makes the choice.
Who maintains them?
Firecrawl itself. The skills live in the vendor's main repository and the `firecrawl/skills` catalog is updated from it automatically.
Discussion
No questions or reviews yet.
Related
SkillsFirecrawl skills catalog for Claude Code, Cursor and Codex: web search, scraping, crawling, page interaction and API integration
Ready-made Firecrawl workflows: sourced research, SEO audits, site QA, knowledge bases and company contact lists
Five skills for the XCrawl API: scraping a single page, mapping a site's URLs, crawling a whole site and search, with cURL and Node examples
The Firecrawl command-line client with agent skills: web search, scraping, crawling, URL mapping and page interaction
The official Firecrawl plugin for finding sources, reading web pages and researching companies in Claude Code
Portable skill for the Crawl4AI CLI and SDK: JavaScript-heavy pages, structured extraction by CSS schemas and batch crawling.
Comparison with similar tools
Cards that share the main search query, and how they differ.
| Tool | Official | Authentication | Repository stars | Main difference |
|---|---|---|---|---|
| Firecrawl Build Skills (this card) | no | 201,311 | ||
| Firecrawl CLI | no | 684 | The CLI gives an agent web data in its session; Build Skills teach writing Firecrawl calls into application code | |
| Firecrawl MCP Server | yes | API key | 7,809 | The MCP server exposes Firecrawl tools to the agent directly; here are instructions for code the agent writes |
README in short
The monorepo README describes the Firecrawl platform and SDKs. Its skills section points to the `firecrawl/skills` catalog and explains that Build skills are authored in the `skills/` folder of this repository, while CLI skills and workflows live in separate repositories.
README badge
Original SKILL.md
---
name: firecrawl-build
description: Integrate Firecrawl into application code whenever a product, agent, or workflow needs web data inside the app: web search, live search results, page scraping, structured extraction, or browser interaction. Do not use this skill for one-off terminal-only web tasks during the current session; use firecrawl/cli for those.
license: ISC
metadata:
author: firecrawl
homepage: https://www.firecrawl.dev
inputs:
- name: FIRECRAWL_API_KEY
required: true
- name: FIRECRAWL_API_URL
required: false
---