How the numbers are measured

What goes in the conformance suite, why the classification unit is the trigger-context pair, and what the capability families are.

The numbers on fidelic.dev are produced by commands. Each number, with the command that produced it, is on the receipts page. This page says how they are built.

What goes in the conformance suite

Every test cites Salesforce’s own documentation or an observation captured from a real org. A test with no source does not go in.

Tests that need a real org are skipped rather than counted, and tests awaiting a real-org capture are marked pending. Both are excluded from the fraction on either side, so the denominator is the set of tests that can run against a stock emulator.

Why the unit is the pair, not the trigger

A trigger declaring seven timings and a one-line before-insert trigger are not the same amount of an org. Counting triggers makes the figure swing on which size of trigger happens to be blocked. The published figure counts trigger-context pairs instead.

On the NPSP corpus the two framings differ. Counting pairs gives 28%. Counting triggers gives 30%. The lower figure is the one published.

Re-check both from one command:

docker run --rm -v "$PWD/mdapi":/mdapi:ro \
  --entrypoint /sfapex fidelic/emulator coverage /mdapi

The output states the pair count and the trigger count on the same line.

What the capability families are

The families listed on the receipts page are internal tier boundaries. They are not Salesforce’s vocabulary. The grouped table on the capability page is folded from them, which is why the wording differs between the two pages.