Benchmarks & tests
What counts as a measurement, what does not, and the conditions a score has to carry before it is published.
Written and covered by tests — not yet observed in a live run
- Why this site did not import 0.0.32Nine proof banks were sealed, their digests matched, and the import was still refused. A capture can be exactly the file that was sealed and still show a product that has moved on.2026-09-13
- How Finexia defines a valid benchmarkEleven conditions, written before we have a single score, so that no score can be published without them.2026-09-10
- A green test is not evidenceTested, packaged and observed are three different claims. Most arguments about what a system can do come from treating them as one.2026-09-10
More in Blog