Turn a site crawl export into a clean text catalog that highlights the best pages for AI and SEO work. Ideal for content and SEO teams that want a fast way to prepare curated links with titles and short descriptions.

The flow starts with a simple form where you enter the site name, a short summary, and upload a CSV from your crawler. The file is parsed, then mapped to seven key fields like URL, title, description, status, indexability, content type, and word count. A filter keeps only pages that return 200, are indexable, and are text HTML. You can also enable an AI step with OpenAI to classify pages as useful content or other content. Each page is formatted as a simple line, then all lines are combined and saved as a downloadable text file. You can swap the last node to upload the file to cloud storage.

Use a CSV export that includes internal URLs, ideally the internal HTML version. The mapping handles multiple languages, so non English exports still work. Expect big time savings by moving from manual sorting to a guided flow. Teams can build a clean list in minutes and reuse the same steps for new sites or larger crawls.

• Form upload that collects site name, short summary, and a CSV file
• CSV extraction that prepares rows for easy processing
• Multi language column mapping for URL, title, description, status, indexability, content type, and word count
• Rules that keep only 200 status, indexable, and text HTML pages
• Optional OpenAI powered classification to flag useful content
• Row builder that formats each page as a simple list line
• Concatenation that combines all lines into one clean text block
• File creation that lets you download the final text file in n8n
• Swap the final step to upload the file to cloud storage if needed

• Reduce manual review from 3 hours to 10 minutes for a 5000 URL crawl
• Improve URL selection accuracy by enforcing 200 status and indexable rules
• Handle 10 times more pages by automating parsing, filtering, and formatting
• Use OpenAI to prioritize high value pages without extra steps
• Create a single text file that any team can read and share

1. Import the template into n8n: Create a new workflow in n8n > Click the three dots menu > Select 'Import from File' > Choose the downloaded JSON file.
2. You'll need accounts with OpenAI. See the Tools Required section above for links to create accounts with these services.
3. In the n8n credentials manager, create an OpenAI credential: Click New > Search for OpenAI > Choose API Key > Name the credential clearly (for example, OpenAI SEO Classifier) > Paste your API key from your OpenAI account > Save.
4. Open the Form trigger node and confirm the three fields are present: site name, site description, and file upload. Copy the form URL so you can submit a test later.
5. Export a CSV of internal pages from your crawler. The internal HTML version works best. The internal all version is supported because the flow filters pages after import.
6. Submit the form with the CSV, a site name, and a short summary. Run the workflow once to create sample data for testing.
7. Open the Extract node output and check that rows are parsed. If fields look empty, confirm the CSV headers match the expected columns in your language.
8. Review the Set useful fields node. Confirm URL, title, description, status, indexability, content type, and word count map correctly. Adjust mappings if your headers differ.
9. Open the Filter URLs node and keep the default rules for 200 status, indexable, and text HTML. Add extra filters like word count or path rules if you want a tighter list.
10. Optional: Enable the Text Classifier node. In its credential dropdown, pick your OpenAI credential. Start with a small batch to check token usage and classification results.
11. Check the Set Row and Concatenate nodes. Confirm the output shows one line per page with the title, link, and description when available.
12. Open the File creation node and run the workflow. Download the text file and review the content. If needed, rename the file to include the site name.
13. Optional: Replace the placeholder upload node with a Google Drive or OneDrive node and configure folder, file name, and upload options.
14. Troubleshooting: If the file fails to parse, verify the CSV delimiter and header names. If classification times out on large sites, add a Loop Over Items node and process in batches. If content type values differ, update the filter to match your crawler output.