Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use when reviewing CI coverage, automated checks, or test strategy related to Maintain test coverage thresholds. Focus on whether the rule is continuously verified, not just documented.
.claude/skills/thedaviddias-test-coverage/SKILL.md| Model | Eval pass | Runs |
|---|---|---|
| gemini-3.6-flash | 100% | 27 |
| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-18 | ✗→✓ | ▲ Improved | -13% | 0% |
| case-01 | ✗→✓ | ▲ Improved | -9% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -2% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -4% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -23% | 0% |
Coverage thresholds prevent code quality from degrading over time by failing builds when test coverage drops below safe levels.
Review this project's test coverage configuration to ensure minimum thresholds are set and enforced in CI/CD.
Configure coverage thresholds in Jest, Vitest, or your testing framework, and add coverage checks to CI/CD.
Explain how to set up test coverage thresholds and what coverage levels are appropriate for different types of code.
Review tests, CI workflows, and enforcement points related to Maintain test coverage thresholds. Flag exact gaps where the rule is not automatically verified or where failures do not block regressions.
For full implementation details, code examples, and framework-specific guidance, see references/rule.md.
Rule page: https://frontendchecklist.io/en/rules/testing/test-coverage
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-18 | fail→pass | 12,647 | 8,591 | -32% | 1 | 1 | 0% | 2,102 | 1,828 | -13% | 0 | 0 | — |
case-19 | pass→pass | 11,602 | 7,573 | -35% | 1 | 1 | 0% | 1,781 | 1,574 | -12% | 0 | 0 | — |
case-01 | fail→pass | 10,619 | 8,497 | -20% | 1 | 1 | 0% | 2,101 | 1,906 | -9% | 0 | 0 | — |
case-02 | fail→pass | 15,689 | 12,977 | -17% | 1 | 1 | 0% | 2,664 | 2,623 | -2% | 0 | 0 | — |
case-03 | fail→pass | 10,409 | 8,017 | -23% | 1 | 1 | 0% | 1,859 | 1,789 | -4% | 0 | 0 | — |
case-04 | pass→pass | 9,901 | 7,797 | -21% | 1 | 1 | 0% | 1,645 | 1,721 | +5% | 0 | 0 | — |
case-05 | pass→pass | 12,802 | 7,359 | -43% | 1 | 1 | 0% | 2,173 | 1,623 | -25% | 0 | 0 | — |
case-06 | pass→pass | 11,096 | 9,288 | -16% | 1 | 1 | 0% | 2,118 | 1,978 | -7% | 0 | 0 | — |
case-07 | pass→pass | 9,029 | 4,534 | -50% | 1 | 1 | 0% | 1,522 | 1,063 | -30% | 0 | 0 | — |
case-08 | pass→pass | 8,448 | 4,977 | -41% | 1 | 1 | 0% | 1,328 | 1,165 | -12% | 0 | 0 | — |
case-09 | fail→pass | 6,480 | 4,313 | -33% | 1 | 1 | 0% | 1,375 | 1,061 | -23% | 0 | 0 | — |
case-10 | pass→pass | 10,439 | 8,985 | -14% | 1 | 1 | 0% | 1,813 | 1,848 | +2% | 0 | 0 | — |
case-11 | pass→pass | 9,383 | 3,920 | -58% | 1 | 1 | 0% | 1,489 | 915 | -39% | 0 | 0 | — |
case-12 | pass→pass | 12,248 | 4,598 | -62% | 1 | 1 | 0% | 2,103 | 1,040 | -51% | 0 | 0 | — |
case-17 | fail→pass | 12,497 | 9,925 | -21% | 1 | 1 | 0% | 2,228 | 2,164 | -3% | 0 | 0 | — |
case-13 | pass→pass | 11,679 | 9,499 | -19% | 1 | 1 | 0% | 2,066 | 2,153 | +4% | 0 | 0 | — |
case-14 | fail→pass | 11,015 | 5,660 | -49% | 1 | 1 | 0% | 1,714 | 1,163 | -32% | 0 | 0 | — |
case-15 | fail→pass | 12,084 | 4,301 | -64% | 1 | 1 | 0% | 2,138 | 1,112 | -48% | 0 | 0 | — |
case-16 | pass→pass | 7,716 | 6,409 | -17% | 1 | 1 | 0% | 1,198 | 1,260 | +5% | 0 | 0 | — |
case-20 | pass→pass | 8,317 | 7,163 | -14% | 1 | 1 | 0% | 1,835 | 1,701 | -7% | 0 | 0 | — |
case-21 | pass→pass | 11,958 | 7,893 | -34% | 1 | 1 | 0% | 2,079 | 1,609 | -23% | 0 | 0 | — |
case-22 | pass→pass | 14,494 | 13,887 | -4% | 1 | 1 | 0% | 2,379 | 2,531 | +6% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +36 percentage points is the difference between those two pass rates over the 22 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.