$ paw-test compare alpha.paw beta.paw suite.yaml --no-fuzz
exit code: 0
--- stdout ---
Differences (1/2):

  Input: today
    alpha -> 2026-09-11
    beta -> tomorrow
    pass differs: alpha=True beta=False
    expected differs: want 2026-09-11 -- alpha=True beta=False

Correct against expected: alpha 2/2 (100.0%) -- all 2 scored; beta 1/2 (50.0%) -- all 2 scored
  1 case(s) where exactly one adapter matched the answer key -- listed under Differences above.

Summary: 2 cases, 1 identical output (byte-for-byte), 1 equivalent output (byte-for-byte + JSON/whitespace-normalized),
alpha pass 2/2, beta pass 1/2, only-alpha-pass 1, only-beta-pass 0, errored 0/2
