# Extract Web Data with Firecrawl

Web content is often hard to collect in a clean format for AI workflows. This skill uses Firecrawl to scrape, crawl, map, search, and extract structured website data.

## Install

```bash
npx skillstore add enactprotocol/enact-firecrawl
```

## Metadata

- Status: approved
- Slug: enactprotocol-enact-firecrawl
- Version: 1.2.1
- Author version: 1.2.1
- Skillstore revision: r1
- Version status: valid
- Tree hash: fe1360523678784aa75adc973b6d15e181454b598e2e3927216aa5529a190b3b
- Author: EnactProtocol
- GitHub username: EnactProtocol
- License: MIT
- Repository: https://github.com/EnactProtocol/enact/tree/main/examples/tools/firecrawl
- Ref: 1ffa7643651792ccb4bd3b15d924d2c97edff755
- Supported tools: Claude, Codex, Claude Code
- Audit status: complete
- Agent install advisory: confirmation\_required
- Manual install advisory: allowed
- Artifact signature: available
- Audit attestation: unavailable
- Human verification: not\_verified
- Risk factors: network, env\_access, external\_commands
- Quality score: 38
- Quality tier: warning
- Public page: https://skillstore.pages.dev/skills/enactprotocol-enact-firecrawl
- Manifest: https://skillstore.pages.dev/api/skills/enactprotocol-enact-firecrawl/manifest

## Capabilities

- Scrapes a single URL and returns Firecrawl content data.
- Crawls a site with a configurable page limit.
- Maps URLs from a website without scraping every page.
- Searches the web through the Firecrawl search endpoint.
- Extracts structured data from pages using a prompt and optional JSON schema.
- Returns operation status, source URL or query, data, and error details.

## Use Cases

- Collect documentation for AI analysis: Crawl a documentation site and convert selected pages into markdown for review or summarization.
- Monitor public web information: Search the web or scrape target pages to gather clean content for research notes.
- Extract structured page data: Use a natural language prompt and optional schema to pull pricing, headlines, links, or metadata from pages.

## Prompt Templates

### Scrape one page

```
Use Firecrawl to scrape this public URL as markdown: [URL]. Return the main content and page metadata.
```

### Crawl a small documentation site

```
Use Firecrawl to crawl [DOCUMENTATION_URL] with a limit of 10 pages. Summarize the key topics found.
```

### Find relevant pages before scraping

```
Use Firecrawl map on [SITE_URL] and filter for pages related to [TOPIC]. List the most relevant URLs.
```

### Extract schema-based information

```
Use Firecrawl extract on [URL]. Extract [FIELDS] using a JSON schema, then explain any missing values.
```

## Limitations

- Requires a valid FIRECRAWL\_API\_KEY secret before use.
- Sends target URLs, prompts, schemas, and requested content to the Firecrawl API.
- Crawl and extract jobs can time out after the configured polling window.
- Results depend on Firecrawl access, site permissions, and the target website response.

## Best Practices

- Use the smallest crawl limit that can answer the task.
- Scrape only public pages that you are authorized to process.
- Store FIRECRAWL\_API\_KEY as a secret and rotate it if exposed.

## Anti Patterns

- Do not scrape private, sensitive, or access-controlled content without authorization.
- Do not request screenshots unless visual page content is required.
- Do not send secrets, credentials, or private data inside extraction prompts or schemas.

## Security Audit

- Audited at: 2026-07-09T04:10:14.642\+00:00
- Summary: The skill is a transparent Firecrawl API wrapper with expected network access and a required API-key secret. Most high static key findings are false positives from legitimate secret handling, but screenshot capture and templated command argument interpolation create privacy and injection concerns that should be addressed before broad marketplace publication.

## Stats

- Views: 138
- Downloads: 12
- Favorites: 0
- Popularity score: 0