Tool Integrations
⚙Tool
Firecrawl
Crawl, scrape, and extract structured data from websites
Firecrawl
Firecrawl crawls websites and extracts clean, structured content. It handles JavaScript rendering, pagination, and complex site structures — returning clean markdown or structured data.
Overview
| Property | Value |
|---|---|
| Type | firecrawl |
| Category | Tool — Web Scraping |
| Auth | API Key |
Operations
| Operation | Description |
|---|---|
| Scrape | Extract content from a single page |
| Crawl | Crawl multiple pages from a starting URL |
| Map | Get a sitemap of all discoverable URLs |
| Extract | Extract structured data using a schema |
Configuration
| Setting | Type | Description |
|---|---|---|
| API Key | Password | Firecrawl API key |
| URL | Short input | Target URL to scrape or crawl |
| Formats | Checkbox list | markdown, html, links, screenshot |
| Max Pages | Slider | Maximum pages to crawl (1–100) |
| Only Main Content | Toggle | Exclude headers, footers, sidebars |
Outputs
| Field | Type | Description |
|---|---|---|
content | string | Extracted content (markdown) |
pages | json | Array of crawled page objects |
metadata | json | Page metadata (title, description) |
Example: Competitive Intelligence
Workflow:
[Starter: Competitor URL] → [Firecrawl: Crawl] → [Agent: Analyze] → [Notion: Save]Crawl a competitor's website (landing pages, pricing, features), then have an Agent analyze the content and save a competitive analysis report to Notion.
Firecrawl Config:
- URL:
{{starter.input}} - Max Pages:
20 - Only Main Content: ✅
- Formats:
markdown
Tips
- Scrape vs Crawl — Scrape for one page, Crawl for multiple pages from a site
- Markdown format is best for AI processing — clean text without HTML noise
- Only Main Content removes navigation, ads, and boilerplate
- Extract mode — define a JSON schema to get structured data from pages