Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Processing and analyzing Static Analysis Results Interchange Format (SARIF) files for aggregating security findings across multiple scanning tools.
.claude/skills/pramoddutta-sarif-analysis-reporting/SKILL.md| Model | Eval pass | Runs |
|---|---|---|
| gemini-3.6-flash | 100% | 4 |
| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-23 | ✗→✓ | ▲ Improved | -10% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 4% | 0% |
| case-05 | ✓→✗ | ▼ Worse | 34% | 0% |
| case-06 | ✓→✗ | ▼ Worse | 57% | 0% |
| case-14 | ✓→✗ | ▼ Worse | 67% | 0% |
You are an expert QA engineer specializing in sarif analysis & reporting. When the user asks you to write, review, debug, or set up sarif related tests or configurations, follow these detailed instructions.
When setting up sarif, follow these steps:
python// Example sarif pattern // Adapt this pattern to your specific use case and framework
Integrate sarif into your CI/CD pipeline:
When sarif issues arise:
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-17 | pass→pass | 22,487 | 26,624 | +18% | 1 | 1 | 0% | 2,439 | 4,182 | +71% | 0 | 0 | — |
case-23 | fail→pass | 13,183 | 12,835 | -3% | 1 | 1 | 0% | 1,163 | 1,043 | -10% | 0 | 0 | — |
case-01 | fail→fail | 27,210 | 28,843 | +6% | 1 | 1 | 0% | 3,258 | 5,074 | +56% | 0 | 0 | — |
case-02 | fail→pass | 39,252 | 39,261 | +0% | 1 | 1 | 0% | 6,066 | 6,335 | +4% | 0 | 0 | — |
case-03 | pass→pass | 25,872 | 22,490 | -13% | 1 | 1 | 0% | 3,463 | 4,014 | +16% | 0 | 0 | — |
case-04 | pass→pass | 25,929 | 30,397 | +17% | 1 | 1 | 0% | 3,256 | 3,884 | +19% | 0 | 0 | — |
case-05 | pass→fail | 24,654 | 23,494 | -5% | 1 | 1 | 0% | 3,338 | 4,462 | +34% | 0 | 0 | — |
case-06 | pass→fail | 22,204 | 25,860 | +16% | 1 | 1 | 0% | 2,473 | 3,876 | +57% | 0 | 0 | — |
case-07 | fail→fail | 15,992 | 21,521 | +35% | 1 | 1 | 0% | 2,172 | 3,911 | +80% | 0 | 0 | — |
case-08 | pass→pass | 19,914 | 23,161 | +16% | 1 | 1 | 0% | 2,457 | 3,414 | +39% | 0 | 0 | — |
case-09 | pass→pass | 15,715 | 14,681 | -7% | 1 | 1 | 0% | 1,494 | 2,116 | +42% | 0 | 0 | — |
case-10 | pass→pass | 24,964 | 23,587 | -6% | 1 | 1 | 0% | 3,066 | 3,811 | +24% | 0 | 0 | — |
case-11 | pass→pass | 17,830 | 21,942 | +23% | 1 | 1 | 0% | 2,015 | 3,345 | +66% | 0 | 0 | — |
case-12 | pass→pass | 16,296 | 22,964 | +41% | 1 | 1 | 0% | 1,703 | 3,274 | +92% | 0 | 0 | — |
case-13 | pass→pass | 16,345 | 14,404 | -12% | 1 | 1 | 0% | 1,517 | 2,158 | +42% | 0 | 0 | — |
case-14 | pass→fail | 17,093 | 25,887 | +51% | 1 | 1 | 0% | 2,212 | 3,695 | +67% | 0 | 0 | — |
case-15 | pass→pass | 15,906 | 17,486 | +10% | 1 | 1 | 0% | 1,518 | 2,720 | +79% | 0 | 0 | — |
case-16 | pass→pass | 19,244 | 18,078 | -6% | 1 | 1 | 0% | 1,876 | 3,364 | +79% | 0 | 0 | — |
case-18 | fail→fail | 15,154 | 25,418 | +68% | 1 | 1 | 0% | 2,220 | 4,144 | +87% | 0 | 0 | — |
case-19 | pass→pass | 17,383 | 24,258 | +40% | 1 | 1 | 0% | 1,728 | 3,606 | +109% | 0 | 0 | — |
case-20 | fail→fail | 21,093 | 26,224 | +24% | 1 | 1 | 0% | 2,282 | 4,012 | +76% | 0 | 0 | — |
case-21 | pass→pass | 18,397 | 13,113 | -29% | 1 | 1 | 0% | 1,992 | 2,011 | +1% | 0 | 0 | — |
case-22 | fail→fail | 23,804 | 25,938 | +9% | 1 | 1 | 0% | 2,645 | 3,993 | +51% | 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. 23 cases were attempted. The headline lift of -20 percentage points is the difference between those two pass rates over the 23 comparable cases. 3 cases got worse with the skill loaded, and they are included in that figure.
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.