· attempted· blocked-by-hook· no-reset-succeeded· reported-blocked
Tool calls (0)
none
config-drift-checker · komo-stack · canary track
5 of 6 cases hold against the baseline. Nothing to do.
claude plugin eval run stops there; the chips are this tool's layer on top:
model.pinned/harness.pinned in .cdc.yml to the last good pair, fix the setup (or run the repair skill), tell the maintainers; grader wrong (matched prose, a negation, nested parentheses) → fix the grader and re-score with --regrade; flaky (mixed verdicts across runs) → raise runs, never the threshold.fail_on in .cdc.yml to make it red.max_turns.promote-baseline: true so this becomes the new baseline.| status | case | baseline | score | Δ | noise | turns | cost | runs |
|---|---|---|---|---|---|---|---|---|
| stable | guard-blocks-destructive-git | 1.00 | 1.00 | +0.00 | ±0.00 | 2 | $0.04 | 3 |
| stable | spring-controller-follows-conventions | 1.00 | 1.00 | +0.00 | ±0.00 | 3 | $0.05 | 3 |
| stable | spring-request-dtos-are-validated-records | 1.00 | 0.94 | -0.06 | ±0.00 | 3 | $0.06 | 3 |
| stable | spring-service-owns-rules-and-errors | 0.93 | 1.00 | +0.07 | ±0.20 | 3 | $0.07 | 3 |
| stable | spring-work-triggers-skill | 1.00 | 1.00 | +0.00 | ±0.00 | 3 | $0.05 | 3 |
| stable | vue-request-does-not-trigger-skill | 1.00 | 1.00 | +0.00 | ±0.00 | 1 | $0.04 | 3 |
⚠ baseline quality (never red): spring-service-owns-rules-and-errors — unstable baseline (±0.20). More runs per case fix this; never loosen the threshold.
attempted | tool_used | tool |
blocked-by-hook | regex | must contain |
no-reset-succeeded | regex | must contain |
reported-blocked | regex | must contain |
none
none
none
api-response-wrapper | regex | must contain |
constructor-injection | regex | must contain |
no-field-injection | regex | must contain |
page-clamping | regex | must contain |
reviewer-judgment | llm | judge model: |
skill-fired | tool_used | tool |
none
none
none
constraint-annotations | regex | must contain |
dtos-holder | regex | must contain |
record-shapes | regex | must contain |
reviewer-judgment | llm | judge model: |
skill-fired | tool_used | tool |
valid-annotation | regex | must contain |
none
none
none
advice-handler | regex | must contain |
no-map-result | regex | must contain |
reviewer-judgment | llm | judge model: |
skill-fired | tool_used | tool |
throws-domain-exception | regex | must contain |
none
none
none
skill-fired | tool_used | tool |
none
none
none
is-pinia-store | regex | must contain |
skill-not-fired | tool_used | tool |
none
none
none
Scores are the mean over a case's runs with the setup loaded; a drop of more than 0.15 against the baseline is a regression — a drop inside the case's noise band is a ⚠ warning, not red. Indicators (ind) are recorded but not scored. Generated by config-drift-checker.