the sample output in the root README:
18 / 100 Unverified · driver unchecked
fingerprint −0 · network −12 · automation −60
the listed checks match the buckets: webdriver −40 + worker −20 = automation −60, and IP timezone −12 = network −12. so 100 − 72 = 28, but the headline says 18.
where do the other 10 points go? is there a hidden deduction or cap, maybe tied to "Unverified"? or is the score not a straight subtraction (weighted, or multiplicative per bucket)? or is the sample just stale?
this matters for --min-score 60. i can't tell how close a profile is to the gate from the per-check numbers if the total isn't their sum. verdict.ts is the source of truth, but a one-line formula in packages/cli/README.md would save reverse engineering it.
the sample output in the root README:
the listed checks match the buckets: webdriver −40 + worker −20 = automation −60, and IP timezone −12 = network −12. so 100 − 72 = 28, but the headline says 18.
where do the other 10 points go? is there a hidden deduction or cap, maybe tied to "Unverified"? or is the score not a straight subtraction (weighted, or multiplicative per bucket)? or is the sample just stale?
this matters for
--min-score 60. i can't tell how close a profile is to the gate from the per-check numbers if the total isn't their sum.verdict.tsis the source of truth, but a one-line formula in packages/cli/README.md would save reverse engineering it.