  !! NOT ENOUGH COVERAGE TO JUDGE - not a regression  (exit 4)
  offline · <run-id> · 2 cases x 2 runs · replay · seed 42
  gating on the Wilson lower bound

!!    fenced-json      coverage    2/4 execution(s) took paths the cassettes do not cover (status_other) - run
                                   `slmeval record --fill-gaps`
XX    fenced-json      pass_k      Wilson lower bound 20.7% against a floor of 90% (point estimate 100.0% over
                                   1 case(s), 2 execution(s))
!!    sloppy-router    coverage    2/4 execution(s) took paths the cassettes do not cover (status_other) - run
                                   `slmeval record --fill-gaps`
XX    sloppy-router    pass_k      Wilson lower bound 0.0% against a floor of 90% (point estimate 0.0% over 1
                                   case(s), 2 execution(s))
XX    prose-only       pass_k      Wilson lower bound 0.0% against a floor of 90% (point estimate 0.0% over 2
                                   case(s), 4 execution(s))

  -- not checked - the figure was never measured, so these neither passed nor failed:
--    fenced-json      cost_ratio    no cost per case to compare: the baseline's was zero
--    fenced-json      p95_ratio     latency was not measured in this run - it ran concurrently, so a p95
                                     would describe the scheduler. Re-run with --serial-latency to enforce
                                     this
--    sloppy-router    cost_ratio    no cost per case to compare: the baseline's was zero
--    sloppy-router    p95_ratio     latency was not measured in this run - it ran concurrently, so a p95
                                     would describe the scheduler. Re-run with --serial-latency to enforce
                                     this
--    prose-only       cost_ratio    no cost per case to compare: the baseline's was zero
--    prose-only       p95_ratio     latency was not measured in this run - it ran concurrently, so a p95
                                     would describe the scheduler. Re-run with --serial-latency to enforce
                                     this

  8/13 check(s) passed, 6 not measured
  -> exit codes: 0 pass · 1 gate failed · 2 usage · 3 environment · 4 coverage
