CrawlVolt
<- Field Notes
Guide / n8n

Extract website Markdown in n8n.

Use a built-in HTTP Request node, inspect one page and connect the Markdown to your existing workflow.

By CrawlVoltUpdated

Prepare a key and a public page.

First, create your API key. Open API keys, enter a name such as markdown-guide and select Create key. Copy the secret when it appears; it is shown once.

Open API keysCreate an account

Already registered? Sign in and continue to API keys.

Store the key in an n8n Header Auth credential, as described below. Choose a public page you own or have permission to collect, then check your available credits.

Download the manual n8n workflow. It is inactive and contains no API key. Import it, then select your own Header Auth credential before running it.

Connect a Manual Trigger to HTTP Request.

  1. Set the method to POST and the URL to https://www.crawlvolt.com/v1/scrape.
  2. Select Generic Credential Type and Header Auth. In the credential, set the header name to Authorization and its value to Bearer YOUR_API_KEY.
  3. Enable Send Body, choose JSON, then enter the body below.
  4. Enable Send Headers and add an Idempotency-Key using the expression below.
JSON
{
  "url": "https://example.com",
  "formats": ["markdown"]
}
n8n expression
{{ $execution.id }}-scrape-page

This example handles one page per execution. Multiple input items need a different stable key for each operation. Keep the response format on JSON and default HTTP error handling enabled.

These settings use the built-in HTTP Request node.

Run once and inspect the result.

With the default body-only response, the Markdown is in outputs.markdown. Reference it in the next node with:

n8n expression
{{ $json.outputs.markdown }}

Check that the required content is present, then connect your storage or indexing step. If the request fails, inspect its status and Activity entry before retrying. A new workflow execution is a new operation.

Choose a refresh frequency after the first useful result.

Keep the workflow manual until the output, credentials and credits are checked. Add scheduling only when you know how often the application needs fresh content and which account limits apply.

See current plans, try a public page or use the Python example.