Skip to content

Fit check

Terminal window
analog assess https://example.com/catalog

The fit check runs locally — no extraction, no backend call, no sign-in — and reports measured facts plus task-keyed guidance:

  • The page’s markdown size, measured (what reading it directly would cost you in tokens).
  • A task menu: working with the page’s data → analog; one quick question, once → read the page directly; coming back repeatedly → extraction pays for itself on the second visit.
  • The page’s declared navigation. Top-level labeled destinations, read locally from the already-fetched HTML — on a landing page, “where can I go from here?” answered before you spend an extraction.
  • Platform data feeds, verified. If the page belongs to a platform that serves its own feed (a Shopify store’s /products.json, a hosted Greenhouse board’s jobs feed), Analog verifies it with one robots-respecting request and reports what it returned just now — stating the feed’s existence, never its authority; a platform-controlled feed can skew, and the rendered page stays your check against it. Skip the request with --no-probe.
  • Sitemaps, on any site. When the site’s robots.txt lists a sitemap, the assessment names it — a machine-readable index of the site’s pages, useful when your task needs many pages rather than this one. Listed is all that’s claimed (the sitemap itself is never fetched), and reading the listing costs nothing beyond the cached robots.txt read. Also skipped by --no-probe.
  • Pagination facts — whether this URL is one page of a larger collection, and how far its own links reach.

Analog does not estimate other tools’ costs or claim a savings figure. The numbers shown are measured; the task-shaped judgment is yours.