the sample block in the root README:
18 / 100 Unverified · driver unchecked
fingerprint −0 · network −12 · automation −60
...
22 checks · 2 critical · 1 warnings · 18 clean
two things don't reconcile:
- 100 − 0 − 12 − 60 = 28, but the score is 18. is there a separate deduction for the
Unverified · driver unchecked state that isn't shown in the three category buckets? if so it's worth naming, since --min-score 60 in CI depends on knowing what can silently take 10 points off.
22 checks but 2 + 1 + 18 = 21. and separately, the run clearly had edge data (edge: 203.0.113.7 · AS4058 … · HTTP/2 · TLSv1.3), so per "40 checks. 32 need nothing but the browser (--offline), 8 compare the JS story" i'd expect 40 here, not 22.
is the count only checks that could be evaluated (probes that returned data), and does the /100 denominator shrink with it, or do unevaluated checks just score as clean? that changes whether comparing two --json runs across chrome versions is meaningful.
the sample block in the root README:
two things don't reconcile:
Unverified · driver uncheckedstate that isn't shown in the three category buckets? if so it's worth naming, since--min-score 60in CI depends on knowing what can silently take 10 points off.22 checksbut 2 + 1 + 18 = 21. and separately, the run clearly had edge data (edge: 203.0.113.7 · AS4058 … · HTTP/2 · TLSv1.3), so per "40 checks. 32 need nothing but the browser (--offline), 8 compare the JS story" i'd expect 40 here, not 22.is the count only checks that could be evaluated (probes that returned data), and does the /100 denominator shrink with it, or do unevaluated checks just score as clean? that changes whether comparing two
--jsonruns across chrome versions is meaningful.