AFP / PDF Template Regression Testing
Catch template drift and structural regressions across production runs before they ship.
FAQ
How is the diff deterministic?
Categorization is a pure function of the two documents’ structured inventories and your tolerances — identical inputs always produce identical findings, so a result is reproducible and auditable.
What counts as a critical regression?
Layout breaks, missing mandatory disclosure blocks, IMb encoding changes, rendering-visible font/image substitution, and page-count changes beyond tolerance — none of which can be suppressed by widening a tolerance.
Can I tune what counts as drift?
Yes — page-count and resource tolerances are configurable per workspace and per template family, with sensible defaults out of the box.
Is this a visual (pixel) diff?
No — it is a structural diff over the document’s parsed inventory. Rendered pages are shown only as supplementary side-by-side evidence, never as the source of a finding.