Sato Score

The Sato Score sums six evidence-only components: maintenance and liveness (observed commits, releases and posts), code transparency, interface verifiability (a probe that answers), docs and demo, listing transparency and provenance, and independent verification. Each has a fixed maximum; the weights and the rules for every point are published on the methodology page. Tiers are High at 70 and above, Medium at 40, Low below that, and Provisional while there is too little history to score.

Only products are scored — research papers, newsletters and communities carry no score, because a score measures a running thing. The score is recomputed daily and its history is snapshotted, so a listing's score over time is a series, not a claim.

What the score is not: it does not say a project is safe, well-built or honest, and it says nothing about returns. A private, new, well-run project scores low because there is less to verify; an abandoned open-source repo can score in the middle on transparency alone. Read the components, not just the number.