Scrape, crawl, map, search, and extract data from websites using Firecrawl. Turn entire websites into LLM-ready markdown or structured data. Supports all Firecrawl API v2 operations including batch processing, team usage tracking, and AI-powered data extraction.
Install and configure the official Firecrawl nodes for n8n to scrape, crawl, map, search, and extract data from websites. Turn entire websites into LLM-ready markdown or structured data with a single API.
This skill guides you through installing and configuring the `@mendable/n8n-nodes-firecrawl` community package in n8n, enabling you to:
1. In your n8n instance, navigate to **Settings** → **Community Nodes**
2. Click **Install a community node**
3. Enter the package name: `@mendable/n8n-nodes-firecrawl`
4. Click **Install**
5. Wait for installation to complete
**For Docker/self-hosted installations:**
```yaml
version: '3'
services:
n8n:
image: n8nio/n8n
environment:
- N8N_COMMUNITY_PACKAGES_ALLOW_TOOL_USAGE=true
ports:
- "5678:5678"
volumes:
- ~/.n8n:/home/node/.n8n
```
Set the environment variable `N8N_COMMUNITY_PACKAGES_ALLOW_TOOL_USAGE=true` to enable AI tool usage.
1. In your n8n workflow, add a **Firecrawl** node
2. When prompted for credentials, click **Create New Credentials**
3. Enter your Firecrawl API key (obtain from https://firecrawl.dev dashboard)
4. Save the credentials
**Security Note:** Keep your API key secure and never commit it to version control.
The Firecrawl node supports the following operations:
#### Core Operations
#### Batch Operations
#### Management Operations
#### Usage Tracking
Each operation has specific parameters. Common configurations:
**For Scrape operation:**
**For Crawl operation:**
**For Map operation:**
**For Extract Data operation:**
To use Firecrawl as a tool for AI agents:
1. Add an **AI Agent** node to your workflow
2. Connect the **Firecrawl** node as a tool input to the AI Agent node
3. The AI agent will automatically decide when to use Firecrawl operations
**Dynamic Parameter Definition:**
Use the `$fromAI()` function to let the AI agent decide parameter values:
```javascript
// Let AI decide the URL to scrape
{{ $fromAI("url", "The URL to scrape", "string") }}
// Let AI decide the search query
{{ $fromAI("query", "Search query for finding relevant pages", "string") }}
// Let AI decide whether to include subdomains
{{ $fromAI("includeSubdomains", "Whether to include subdomains", "boolean", false) }}
```
Click the **"Let the model define this parameter"** button next to any field to enable this feature.
1. Click **Execute Node** to test the Firecrawl operation
2. Review the output in the node's output panel
3. Verify the data format matches your expectations
4. Adjust parameters as needed
1. **Start with Map**: For unknown websites, use the Map operation first to discover available URLs
2. **Set reasonable limits**: Configure crawl depth and page limits to prevent excessive API usage
3. **Use Include/Exclude patterns**: Filter URLs efficiently with path patterns instead of crawling everything
4. **Handle errors gracefully**: Add error handling nodes to manage failed scrapes or crawls
5. **Monitor usage**: Regularly check your token and credit usage with the tracking operations
6. **Cache results**: Store scraped data to avoid re-scraping the same content
7. **Use batch operations**: For multiple URLs, use Batch Scrape instead of individual scrape operations
8. **Provide clear prompts**: When using Extract, write specific prompts and schemas for better AI extraction
1. **Use clear operation names**: AI agents rely on operation descriptions to make decisions
2. **Provide context in prompts**: Give the agent clear instructions on what data to extract
3. **Set default limits**: Configure sensible defaults to prevent runaway API costs
4. **Chain operations**: Use Map → Scrape/Crawl workflows for better results
5. **Test incrementally**: Start with small limits when testing AI agent integrations
Leave a review
No reviews yet. Be the first to review this skill!
# Download SKILL.md from killerskills.ai/api/skills/firecrawl-n8n-integration/raw