bot/skills/hanzo-crawl/SKILL.md

name, description, metadata

name	description	metadata
hanzo-crawl	Crawl and scrape web pages via the Hanzo Crawl API (Crawl4AI). Extract structured content and markdown from URLs, preview pages, and crawl sites with JavaScript rendering support. Use when a bot needs to scrape a webpage, crawl a site, or extract content from URLs.	bot requires primaryEnv emoji bins python3 HANZO_API_KEY 🕷

Hanzo Crawl -- Web Scraping API (Crawl4AI)

Crawl web pages and extract structured markdown content via Crawl4AI. Supports JavaScript-rendered pages, CSS selectors for targeted extraction, and async crawl processing.

API Endpoints

Base URL: https://crawl.hanzo.ai

Endpoint	Method	Purpose
/crawl	POST	Crawl a URL and extract content
/md	POST	Quick markdown extraction from a URL
/task/{id}	GET	Poll async crawl task status
/health	GET	Health check

Authentication

All requests require a Bearer token in the Authorization header. Use the HANZO_API_KEY env var.

Authorization: Bearer <token>

Crawl a URL

python3 {baseDir}/scripts/crawl.py --url "https://docs.example.com"

Request Body (POST /crawl)

{
  "urls": "https://docs.example.com/getting-started",
  "word_count_threshold": 10,
  "css_selector": "main, article",
  "excluded_tags": ["nav", "footer", "aside"],
  "wait_for": ".content-loaded",
  "magic": false,
  "screenshot": false
}

Fields

• urls (required): URL string or array of URLs to crawl
• word_count_threshold (optional): Minimum word count for content blocks (default 10)
• css_selector (optional): CSS selector for targeted content extraction
• excluded_tags (optional): Array of HTML tags to exclude from extraction
• wait_for (optional): CSS selector to wait for before extraction (for JS-rendered pages)
• js_code (optional): Array of JavaScript code strings to execute before extraction
• magic (optional): Enable magic mode for complex pages (default false)
• screenshot (optional): Take a screenshot of the page (default false)

Response

The /crawl endpoint returns a task_id for async processing. Poll /task/{task_id} until completion.

{
  "task_id": "abc123-def456",
  "status": "completed",
  "result": {
    "url": "https://docs.example.com/getting-started",
    "success": true,
    "markdown": "# Getting Started\n\nTo deploy your application...",
    "links": {
      "internal": [{ "href": "/next-steps", "text": "Next Steps" }],
      "external": [{ "href": "https://github.com/...", "text": "Source" }]
    },
    "media": {
      "images": [{ "src": "...", "alt": "..." }]
    }
  }
}

Preview a Page (Markdown)

Quick markdown extraction without async processing.

python3 {baseDir}/scripts/preview.py --url "https://docs.example.com/page"

Request Body (POST /md)

{
  "url": "https://docs.example.com/getting-started"
}

Response

{
  "url": "https://docs.example.com/getting-started",
  "success": true,
  "title": "Getting Started",
  "markdown": "# Getting Started\n\nFull extracted markdown content...",
  "status_code": 200
}

Scripts

scripts/crawl.py

Crawl a URL and extract content via Crawl4AI.

python3 {baseDir}/scripts/crawl.py \
  --url "https://docs.example.com" \
  --css-selector "main, article" \
  --excluded-tags "nav,footer,aside" \
  --token "$HANZO_API_KEY"

scripts/preview.py

Quick markdown preview of a page.

python3 {baseDir}/scripts/preview.py \
  --url "https://docs.example.com/page" \
  --token "$HANZO_API_KEY"

Billing

Crawl operations are billed per page crawled. Preview (markdown) operations are billed at a reduced rate. Usage is tracked automatically through the bot gateway.

Environment Variables

HANZO_API_KEY=...                                  # API token for authentication
HANZO_CRAWL_BASE_URL=https://crawl.hanzo.ai        # Override API base URL