Nobody should maintain a cookie spreadsheet.
Katla crawls your site the way a visitor would, records every cookie it sets, and classifies each one with a confidence score. You review and override; the policy follows automatically.
Four passes, one inventory.
Crawl
A headless browser walks your site from the homepage, following links that stay on your domain and refusing navigations that leave it.
Up to 5,000 pages per scanDetect
Every cookie a page sets is recorded — first-party and third-party alike — with its domain, expiry, flags, and the page it appeared on.
Includes HttpOnly and Secure cookiesClassify
Each cookie is matched against a known-tracker database. Whatever the database cannot answer goes to the model with its full context.
Open Cookie Database + LLM fallbackPublish
Cookie and privacy policies are generated from what the scan actually found, and republished whenever the inventory changes.
13 languages · embed or fetch as markdownEvery call shows its reasoning.
Each cookie is matched against a database of known trackers first. Anything left over is read in context — who sets it, what the value looks like, how long it lives — and the model returns a category, a purpose in plain language, and a confidence score. Every call is yours to override.
Re-scans on your cadence
Monthly, weekly or daily depending on plan, without anyone remembering to press the button.
The last word is yours
Reclassify any cookie and write your own description per language. Your wording is what visitors and your policy see.
Scan from your pipeline
Trigger a scan from the CLI on every deploy, then export the classified inventory as CSV for the audit trail.
See what's actually on your site.
Add a domain, verify it once with a DNS record or a meta tag, and the first scan runs in about a minute. Free plan, no card.