feat: Add Firecrawl Tool Node support for n8n AI Agent system

README.md

@@ -8,6 +8,7 @@ This is an n8n community node. It lets you use **[Firecrawl](https://firecrawl.d
 
 [Installation](#installation)  
 [Operations](#operations)  
+[Tool Node Support](#tool-node-support)  
 [Credentials](#credentials)  
 [Compatibility](#compatibility)  
 [Resources](#resources)  
@@ -76,7 +77,79 @@ The **Firecrawl** node supports the following operations:
 - Get historical token usage for your team
 
 ### Team Queue Status
-- Get your team’s current queue load (waiting, active, max concurrency)
+- Get your team's current queue load (waiting, active, max concurrency)
+
+## Tool Node Support
+
+The Firecrawl node now supports **Tool Node** functionality for n8n's AI Agent system! This makes Firecrawl seamlessly usable by AI Agents in n8n workflows.
+
+### Available Tools
+
+When used as a Tool Node, Firecrawl exposes the following tools that AI Agents can use:
+
+#### 1. **scrape_url**
+Scrapes a URL and extracts its content in LLM-ready format (markdown, HTML, or structured data).
+
+**Parameters:**
+- `url` (required): The URL to scrape
+- `formats`: Output formats (comma-separated): markdown, html, rawHtml, screenshot, summary
+- `onlyMainContent`: Only return main content (default: true)
+- `actions`: JSON array of actions to interact with dynamic content
+- `headers`: JSON object of custom headers
+- `waitFor`: Wait milliseconds for page to load
+
+#### 2. **extract_data**
+Extracts structured data from one or more URLs using AI. Define a schema to get specific data fields from pages.
+
+**Parameters:**
+- `urls` (required): Comma-separated list of URLs to extract from (supports glob patterns)
+- `schema` (required): JSON schema defining the structure of data to extract
+- `prompt`: Optional prompt to guide the extraction
+- `enableWebSearch`: Enable web search to find additional data
+- `ignoreSitemap`: Ignore the website sitemap
+- `includeSubdomains`: Include subdomains
+
+#### 3. **search_website**
+Search through a website and optionally scrape the results.
+
+**Parameters:**
+- `url` (required): The URL to search on
+- `search` (required): Search query
+- `scrape`: Whether to scrape search results
+- `limit`: Maximum number of results
+
+#### 4. **crawl_website**
+Crawls an entire website and returns structured data from all pages.
+
+**Parameters:**
+- `url` (required): The URL to start crawling from
+- `limit`: Maximum number of pages to crawl (default: 100)
+- `excludePaths`: Comma-separated path patterns to exclude
+- `includePaths`: Comma-separated path patterns to include
+- `prompt`: Natural language prompt to guide the crawl
+- `formats`: Output formats
+- `allowExternalLinks`: Allow crawling external domains
+- `allowSubdomains`: Allow crawling subdomains
+
+#### 5. **map_website**
+Get all URLs from a website without scraping content.
+
+**Parameters:**
+- `url` (required): The URL to map
+- `limit`: Maximum number of URLs (default: 1000)
+- `excludePaths`: Path patterns to exclude
+- `includePaths`: Path patterns to include
+
+### Using Firecrawl with n8n AI Agents
+
+To use Firecrawl as a Tool Node in your n8n AI Agent workflows:
+
+1. Add a Firecrawl node to your workflow
+2. Configure your Firecrawl credentials
+3. The AI Agent will automatically discover and use Firecrawl tools
+4. When the agent needs to scrape, extract, search, or crawl, it will use the appropriate Firecrawl tool
+
+This makes it easier to integrate web data extraction into AI Agent workflows without manual configuration.
 
 ## Credentials
 
@@ -103,6 +176,12 @@ To use the Firecrawl node, you need to:
 
 ## Version history
 
+### 1.0.7
+- Add Tool Node support for n8n AI Agent system
+- Firecrawl can now be used as a Tool Node in AI Agent workflows
+- Added tool definitions for scrape_url, extract_data, search_website, crawl_website, and map_website
+- Improved discoverability and easier setup for AI Agent integrations
+
 ### 1.0.6
 - Add support for additional Firecrawl endpoints:
   - Batch Scrape (start/status/errors)

nodes/Firecrawl/Firecrawl.node.ts

@@ -1,12 +1,15 @@
 import type { INodeType, INodeTypeDescription } from 'n8n-workflow';
 import { allMethods } from './methods';
 import { allProperties } from './properties';
+import { tools } from './tools';
+import type { Tool } from './tools';
 
 /**
  * Firecrawl API Node implementation
+ * Supports both standard node and Tool Node formats for n8n AI Agent integration
  */
 export class Firecrawl implements INodeType {
-	description: INodeTypeDescription = {
+	description: INodeTypeDescription & { tools?: Tool[] } = {
 		displayName: 'Firecrawl',
 		name: 'firecrawl',
 		icon: 'file:firecrawl.svg',
@@ -36,6 +39,8 @@ export class Firecrawl implements INodeType {
 			},
 		},
 		properties: allProperties,
+		// Add Tool Node support for n8n AI Agent system
+		tools: tools,
 	};
 
 	methods = allMethods;

nodes/Firecrawl/tools.ts

@@ -0,0 +1,233 @@
+/**
+ * Tool definitions for Firecrawl operations in n8n AI Agent system
+ * These tools enable Firecrawl to function as a Tool Node in n8n
+ * 
+ * Note: The 'tools' property in INodeTypeDescription is currently a newer feature
+ * in n8n. If ITool type is not available, we define the structure manually.
+ */
+export interface ToolParameter {
+	name: string;
+	type: string;
+	required?: boolean;
+	description?: string;
+	placeholder?: string;
+	default?: string | boolean | number;
+}
+
+export interface Tool {
+	name: string;
+	description: string;
+	parameters: ToolParameter[];
+}
+
+export const tools: Tool[] = [
+	{
+		name: 'scrape_url',
+		description: 'Scrapes a URL and extracts its content in LLM-ready format (markdown, HTML, or structured data). Use this to get clean, formatted content from any webpage.',
+		parameters: [
+			{
+				name: 'url',
+				type: 'string',
+				required: true,
+				description: 'The URL to scrape',
+				placeholder: 'https://example.com',
+			},
+			{
+				name: 'formats',
+				type: 'string',
+				description: 'Output formats (comma-separated): markdown, html, rawHtml, screenshot, summary',
+				default: 'markdown',
+			},
+			{
+				name: 'onlyMainContent',
+				type: 'boolean',
+				description: 'Only return main content excluding headers, navs, footers',
+				default: true,
+			},
+			{
+				name: 'actions',
+				type: 'string',
+				description: 'JSON array of actions to interact with dynamic content (click, wait, scroll, etc.)',
+			},
+			{
+				name: 'headers',
+				type: 'string',
+				description: 'JSON object of custom headers to send with the request',
+			},
+			{
+				name: 'waitFor',
+				type: 'number',
+				description: 'Wait milliseconds for page to load before fetching content',
+				default: 0,
+			},
+		],
+	},
+	{
+		name: 'extract_data',
+		description: 'Extracts structured data from one or more URLs using AI. Define a schema to get specific data fields from pages.',
+		parameters: [
+			{
+				name: 'urls',
+				type: 'string',
+				required: true,
+				description: 'Comma-separated list of URLs to extract from (supports glob patterns)',
+				placeholder: 'https://example.com/*',
+			},
+			{
+				name: 'schema',
+				type: 'string',
+				required: true,
+				description: 'JSON schema defining the structure of data to extract',
+				placeholder: '{"title": "string", "price": "number"}',
+			},
+			{
+				name: 'prompt',
+				type: 'string',
+				description: 'Optional prompt to guide the extraction process',
+			},
+			{
+				name: 'enableWebSearch',
+				type: 'boolean',
+				description: 'Enable web search to find additional data',
+				default: false,
+			},
+			{
+				name: 'ignoreSitemap',
+				type: 'boolean',
+				description: 'Ignore the website sitemap when crawling',
+				default: true,
+			},
+			{
+				name: 'includeSubdomains',
+				type: 'boolean',
+				description: 'Include subdomains of the website',
+				default: false,
+			},
+		],
+	},
+	{
+		name: 'search_website',
+		description: 'Search through a website and optionally scrape the results. Useful for finding specific content across a domain.',
+		parameters: [
+			{
+				name: 'url',
+				type: 'string',
+				required: true,
+				description: 'The URL to search on',
+				placeholder: 'https://example.com',
+			},
+			{
+				name: 'search',
+				type: 'string',
+				required: true,
+				description: 'Search query to find content',
+				placeholder: 'keywords to search for',
+			},
+			{
+				name: 'scrape',
+				type: 'boolean',
+				description: 'Whether to scrape the search results',
+				default: false,
+			},
+			{
+				name: 'limit',
+				type: 'number',
+				description: 'Maximum number of search results to return',
+				default: 10,
+			},
+		],
+	},
+	{
+		name: 'crawl_website',
+		description: 'Crawls an entire website and returns structured data from all pages. Use this to get content from multiple pages on a domain.',
+		parameters: [
+			{
+				name: 'url',
+				type: 'string',
+				required: true,
+				description: 'The URL to start crawling from',
+				placeholder: 'https://example.com',
+			},
+			{
+				name: 'limit',
+				type: 'number',
+				description: 'Maximum number of pages to crawl',
+				default: 100,
+			},
+			{
+				name: 'excludePaths',
+				type: 'string',
+				description: 'Comma-separated path patterns to exclude (e.g., "blog/*,admin/*")',
+			},
+			{
+				name: 'includePaths',
+				type: 'string',
+				description: 'Comma-separated path patterns to include',
+			},
+			{
+				name: 'prompt',
+				type: 'string',
+				description: 'Natural language prompt to guide the crawl',
+			},
+			{
+				name: 'formats',
+				type: 'string',
+				description: 'Output formats (comma-separated): markdown, html, screenshot',
+				default: 'markdown',
+			},
+			{
+				name: 'allowExternalLinks',
+				type: 'boolean',
+				description: 'Allow crawling external domains',
+				default: false,
+			},
+			{
+				name: 'allowSubdomains',
+				type: 'boolean',
+				description: 'Allow crawling subdomains',
+				default: false,
+			},
+		],
+	},
+	{
+		name: 'map_website',
+		description: 'Get all URLs from a website without scraping content. Use this to discover all pages on a domain.',
+		parameters: [
+			{
+				name: 'url',
+				type: 'string',
+				required: true,
+				description: 'The URL to map',
+				placeholder: 'https://example.com',
+			},
+			{
+				name: 'limit',
+				type: 'number',
+				description: 'Maximum number of URLs to return',
+				default: 1000,
+			},
+			{
+				name: 'excludePaths',
+				type: 'string',
+				description: 'Comma-separated path patterns to exclude',
+			},
+			{
+				name: 'includePaths',
+				type: 'string',
+				description: 'Comma-separated path patterns to include',
+			},
+		],
+	},
+];
+
+/**
+ * Maps tool names to their corresponding Firecrawl operations
+ */
+export const toolToOperationMap: Record<string, string> = {
+	scrape_url: 'scrape',
+	extract_data: 'extract',
+	search_website: 'search',
+	crawl_website: 'crawl',
+	map_website: 'map',
+};
+