Jetscrape MCP connector

The Jetscrape Node calls JetScrape's core scraping tools through its custom MCP server, authenticated with your JetScrape API key. It covers unblocked page scraping, HTML inspection, AI extraction against a JSON schema, and XHR/fetch endpoint discovery.

Parameters

Required Parameters

• action (string): The target JetScrape MCP tool action to run. Must be one of the enabled actions:
  • jetscrape_scrape_unblocker: Scrape one or more URLs through the main engine (Cloudflare, anti-bot, JS rendering). Inputs: apiKey, url or urls. Optional: selectors, mode, wait_until, actions, options, max_items.
  • jetscrape_get_html: Fetch HTML of one URL, condensed by default, to work out selectors. Inputs: apiKey, url.
  • jetscrape_ai_scrape: AI-driven extraction of a URL against a target JSON schema. Inputs: apiKey, url, schema.
  • jetscrape_discover_api_endpoints: Load a page in a real browser and return a compact index of its XHR/fetch calls. Inputs: apiKey, url.
  • jetscrape_read_captured_response: Full body of one response captured by jetscrape_discover_api_endpoints. Inputs: apiKey, dispatch_id, index.
• apiKey (string): Your JetScrape API key, supplied through the node connection.

Optional Parameters

• params (object): A key-value dictionary containing action-specific inputs (e.g., {'url': 'https://example.com', 'selectors': 'h1'}).

Practical Use Cases

1. Structured extraction: Run jetscrape_ai_scrape with a JSON schema to turn a page into typed records.
2. API reverse engineering: Use jetscrape_discover_api_endpoints, then jetscrape_read_captured_response to read the underlying JSON a site loads.