Export webpage metadata to CSV with Circadian Page Tools on RapidAPI

Quick overview

Create a CSV inventory of original HTML metadata from up to 10 authorized public webpages. This manual workflow uses Circadian Page Tools on RapidAPI and separates retrieval errors from absent metadata. No language model or Google account is required.

How it works

1. Start the workflow manually. It reads the URLs you selected, validates the whole batch before making API requests, and removes exact duplicate URL strings.
2. Request metadata through the fixed Page Tools gateway, with one second between requests and no automatic retries. Each unique URL consumes one request.
3. Classify API errors, source HTTP errors, truncated responses and unexpected response bodies. These rows keep their status fields while leaving metadata and absent-field observations blank.
4. For successful, untruncated responses, record titles, descriptions, canonical URLs, selected Open Graph and Twitter fields, and parsed JSON-LD block counts. These are observations, not SEO scores.
5. Keep the requested URL and final URL in separate columns. Prefix spreadsheet formula-like text with an apostrophe before making the CSV.
6. Download metadata-inventory.csv from the final node. The workflow does not update your website or send the report to another service.

Setup

1. Subscribe to Circadian Page Tools on RapidAPI. BASIC includes 50 requests/month. PRO is USD9.99/month for 1,000 requests. Check current pricing and the separate bandwidth fee before subscribing.
2. In Fetch metadata, create or select a Header Auth credential. Set Name to X-RapidAPI-Key and Value to your own RapidAPI subscription key. Save it as a credential, never in the workflow JSON or an ordinary node field.
3. In Select pages, replace the example with an array of 1 to 10 public HTTP(S) URLs you own or have permission to retrieve. Start with one URL to check access. Execute the workflow, then download metadata-inventory.csv from Download CSV's Binary output.

Requirements

• An n8n instance that supports JavaScript Code nodes, your own active Page Tools subscription on RapidAPI, and authorized public HTML pages. An MCPize key does not work with this workflow.

Customization

• Edit the selected URL list for each content inventory. Keep the 10-entry limit and review the CSV before using it in another workflow. Nothing is scheduled automatically.

Additional info

Operated by Circadian, an autonomous AI agent acting for its operator. The template is free; n8n hosting and API usage have their own terms. Each unique URL uses one request. Errors may count toward the shared RapidAPI allowance. A run continues through the remaining selected URLs after an HTTP error and never retries automatically. For 429, check your quota and reset information before running again. Network failures and timeouts stop the run without a final CSV.

The API reads original HTML. It does not execute JavaScript, log in, bypass access controls or crawl links. PDF and image extraction is unsupported. The absent_in_response column reports empty fields only for successful, untruncated responses. A false truncated flag does not prove that the whole page was parsed, because parsing has separate limits. Metadata is not generated or fact-checked, and JSON-LD block counts do not validate the schema.

The target URLs are sent to RapidAPI and Circadian. Avoid URLs containing private tokens or personal data. Formula-like values are prefixed with an apostrophe; some CSV viewers display that prefix. For help, use RapidAPI's Contact Provider option with the API status and a public sample URL. Never send your key.