Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Interpret datadog-inspector findings and translate Datadog monitoring, audit, log-retention, SSO, and RBAC results into GRC evidence and remediation.
.claude/skills/grcengclub-datadog-inspector-expert/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-02 | ✗→✓ | ▲ Improved | -13% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -26% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -25% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -9% | 0% |
| case-10 | ✗→✓ | ▲ Improved | -69% | 0% |
Use this skill when reviewing datadog-inspector output or planning Datadog control remediation.
Findings are written to:
text~/.cache/claude-grc/findings/datadog-inspector/<run_id>.json
The connector emits a datadog_organization resource with SCF evaluations for:
DCH-07: log retentionLOG-08: audit log availabilityMON-01.2: critical monitor coverageIAC-04: SSO enforcement visibilityIAC-07: RBAC role inventory visibilityGOV-03: organization/API inventoryinconclusive as a permission or plan-coverage gap, not a pass.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 7,084 | 16,124 | +128% | 1 | 1 | 0% | 395 | 591 | +50% | 0 | 0 | — |
case-02 | fail→pass | 20,769 | 14,498 | -30% | 1 | 1 | 0% | 3,056 | 2,665 | -13% | 0 | 0 | — |
case-03 | pass→pass | 8,861 | 9,656 | +9% | 1 | 1 | 0% | 1,543 | 2,113 | +37% | 0 | 0 | — |
case-04 | pass→pass | 18,847 | 14,721 | -22% | 1 | 1 | 0% | 3,785 | 3,114 | -18% | 0 | 0 | — |
case-05 | pass→pass | 12,625 | 13,817 | +9% | 1 | 1 | 0% | 2,385 | 2,849 | +19% | 0 | 0 | — |
case-06 | fail→pass | 13,407 | 7,771 | -42% | 1 | 1 | 0% | 1,917 | 1,426 | -26% | 0 | 0 | — |
case-07 | fail→pass | 12,882 | 7,866 | -39% | 1 | 1 | 0% | 1,856 | 1,396 | -25% | 0 | 0 | — |
case-08 | fail→pass | 12,425 | 9,107 | -27% | 1 | 1 | 0% | 1,835 | 1,664 | -9% | 0 | 0 | — |
case-09 | pass→pass | 14,180 | 1,732 | -88% | 1 | 1 | 0% | 2,180 | 517 | -76% | 0 | 0 | — |
case-10 | fail→pass | 11,379 | 3,381 | -70% | 1 | 1 | 0% | 1,720 | 527 | -69% | 0 | 0 | — |
case-11 | fail→pass | 6,654 | 1,914 | -71% | 1 | 1 | 0% | 1,042 | 496 | -52% | 0 | 0 | — |
case-12 | fail→pass | 12,429 | 2,668 | -79% | 1 | 1 | 0% | 2,078 | 551 | -73% | 0 | 0 | — |
case-13 | fail→pass | 10,362 | 1,983 | -81% | 1 | 1 | 0% | 1,672 | 575 | -66% | 0 | 0 | — |
case-14 | fail→pass | 15,164 | 1,442 | -90% | 1 | 1 | 0% | 2,617 | 475 | -82% | 0 | 0 | — |
case-15 | fail→pass | 10,096 | 2,115 | -79% | 1 | 1 | 0% | 1,552 | 591 | -62% | 0 | 0 | — |
case-16 | fail→pass | 12,474 | 5,940 | -52% | 1 | 1 | 0% | 1,859 | 1,192 | -36% | 0 | 0 | — |
case-17 | pass→pass | 14,228 | 9,237 | -35% | 1 | 1 | 0% | 2,156 | 1,640 | -24% | 0 | 0 | — |
case-18 | pass→pass | 14,216 | 7,448 | -48% | 1 | 1 | 0% | 2,372 | 1,455 | -39% | 0 | 0 | — |
case-19 | pass→pass | 12,319 | 7,046 | -43% | 1 | 1 | 0% | 1,783 | 1,362 | -24% | 0 | 0 | — |
case-20 | pass→pass | 14,120 | 8,446 | -40% | 1 | 1 | 0% | 2,210 | 1,469 | -34% | 0 | 0 | — |
case-21 | fail→pass | 14,919 | 2,611 | -82% | 1 | 1 | 0% | 2,499 | 708 | -72% | 0 | 0 | — |
case-22 | pass→pass | 15,629 | 7,640 | -51% | 1 | 1 | 0% | 2,457 | 1,465 | -40% | 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, and 21 counted toward the lift figure. The other 1 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +55 percentage points is the difference between those two pass rates over the 21 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.