Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use when creating a draw.io diagram for evidence collection, evidence lifecycle, audit evidence pipelines, and systems of record in a GRC, security, audit, compliance, privacy, cloud, or risk context.
.claude/skills/grcengclub-grc-evidence-flow-diagram/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-13 | ✗→✓ | ▲ Improved | 12% | 0% |
| case-19 | ✗→✓ | ▲ Improved | 36% | 0% |
| case-08 | ✓→✗ | ▼ Worse | 26% | 0% |
| case-11 | ✓→✗ | ▼ Worse | 30% | 0% |
| case-15 | ✓→✗ | ▼ Worse | 25% | 0% |
Use this skill to structure the GRC content and visual pattern for evidence collection, evidence lifecycle, audit evidence pipelines, and systems of record. Then use the drawio skill to generate the native editable .drawio file and optional PNG/SVG/PDF export.
Include these when relevant:
Produce a Swimlane evidence lifecycle or evidence data-flow diagram. Choose a layout that matches the audience:
drawio skill.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-09 | fail→fail | 29,765 | 30,253 | +2% | 1 | 1 | 0% | 6,176 | 6,694 | +8% | 0 | 0 | — |
case-01 | fail→fail | 32,469 | 34,651 | +7% | 1 | 1 | 0% | 6,216 | 6,734 | +8% | 0 | 0 | — |
case-02 | fail→fail | 32,468 | 33,169 | +2% | 1 | 1 | 0% | 6,214 | 6,732 | +8% | 0 | 0 | — |
case-03 | fail→fail | 28,488 | 30,618 | +7% | 1 | 1 | 0% | 6,207 | 6,726 | +8% | 0 | 0 | — |
case-04 | fail→fail | 31,755 | 30,588 | -4% | 1 | 1 | 0% | 6,189 | 6,707 | +8% | 0 | 0 | — |
case-05 | fail→fail | 31,693 | 32,499 | +3% | 1 | 1 | 0% | 6,189 | 6,707 | +8% | 0 | 0 | — |
case-06 | fail→fail | 30,158 | 32,451 | +8% | 1 | 1 | 0% | 6,187 | 6,705 | +8% | 0 | 0 | — |
case-07 | pass→pass | 30,266 | 32,532 | +7% | 1 | 1 | 0% | 6,185 | 6,703 | +8% | 0 | 0 | — |
case-08 | pass→fail | 28,148 | 31,614 | +12% | 1 | 1 | 0% | 5,299 | 6,685 | +26% | 0 | 0 | — |
case-10 | fail→fail | 27,762 | 31,626 | +14% | 1 | 1 | 0% | 5,778 | 6,698 | +16% | 0 | 0 | — |
case-11 | pass→fail | 25,843 | 31,720 | +23% | 1 | 1 | 0% | 5,156 | 6,685 | +30% | 0 | 0 | — |
case-12 | pass→pass | 22,940 | 31,071 | +35% | 1 | 1 | 0% | 4,711 | 6,684 | +42% | 0 | 0 | — |
case-13 | fail→pass | 28,674 | 28,939 | +1% | 1 | 1 | 0% | 5,957 | 6,693 | +12% | 0 | 0 | — |
case-14 | pass→pass | 25,848 | 29,968 | +16% | 1 | 1 | 0% | 5,151 | 6,696 | +30% | 0 | 0 | — |
case-15 | pass→fail | 28,933 | 30,901 | +7% | 1 | 1 | 0% | 5,358 | 6,684 | +25% | 0 | 0 | — |
case-16 | pass→fail | 19,287 | 33,684 | +75% | 1 | 1 | 0% | 3,245 | 6,692 | +106% | 0 | 0 | — |
case-17 | pass→pass | 11,907 | 25,014 | +110% | 1 | 1 | 0% | 1,899 | 5,475 | +188% | 0 | 0 | — |
case-18 | pass→pass | 30,357 | 32,004 | +5% | 1 | 1 | 0% | 5,868 | 6,682 | +14% | 0 | 0 | — |
case-19 | fail→pass | 24,988 | 31,575 | +26% | 1 | 1 | 0% | 4,900 | 6,682 | +36% | 0 | 0 | — |
case-20 | fail→fail | 12,954 | 12,683 | -2% | 1 | 1 | 0% | 2,063 | 2,417 | +17% | 0 | 0 | — |
case-21 | pass→pass | 13,223 | 16,234 | +23% | 1 | 1 | 0% | 2,499 | 3,743 | +50% | 0 | 0 | — |
case-22 | pass→pass | 10,127 | 11,402 | +13% | 1 | 1 | 0% | 1,898 | 2,694 | +42% | 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 -25 percentage points is the difference between those two pass rates over the 22 comparable cases. 5 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.