Muse Spark 1.2

Meta

Rank #8 of 11 · 30 PRs · 39/95 golden bugs found

37.8

F1 · +0.3 vs avg

Recall
41.0%+5.8 vs avg
Precision
35.0%-9.4 vs avg
Cost / PR
$0.490
Cost / bug found
$0.38

Confidence

95% bootstrap interval on recall (2000 resamples over the 30 PRs):31.3–52.0points. This measures sampling variance from which PRs are in the set.

Judge run-to-run variance not yet measured for this model.

Neither measures model run-to-run variance — re-running the review agent itself, not just the judge. One pass per entry.

Run configuration

HarnesskodusAccess pathapiExecution modereplayReasoningvendor-defaultJudgeclaude-haiku-4-5

Severity mix

What the model called its own findings — not recall by severity, goldens aren't severity-tagged.

Low 2Medium 55High 49Critical 4

Category mix

Only bug/performance/security are consistent across models — the rest is free text, bucketed as other.

Bug110Performance0Security0Other7

By repository

RepoRecallPrecisionGoldensPRs
cal.com39.1%45.0%236
Discourse50.0%49.9%206
Sentry36.8%39.4%196
Keycloak11.8%20.0%176
Grafana68.8%49.2%166

Per-PR breakdown (30)