Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Interpret drata-inspector findings generated from drata-cli workflows and turn Drata control, monitor, evidence, personnel, and integration posture into GRC action.
.claude/skills/grcengclub-drata-inspector-expert/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-06 | ✗→✓ | ▲ Improved | -75% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -76% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -68% | 0% |
| case-10 | ✗→✓ | ▲ Improved | -50% | 0% |
| case-11 | ✗→✓ | ▲ Improved | -46% | 0% |
Use this skill when reviewing drata-inspector output or planning remediation from Drata workflow findings.
drata-inspector wraps the MIT-licensed drata-cli workflow commands. It does not reimplement Drata APIs and does not require Drata MCP.
Findings are written to:
text~/.cache/claude-grc/findings/drata-inspector/<run_id>.json
Resource types:
drata_tenant: summary status across controls, monitors, personnel, and integrationsdrata_control: failing or incomplete controls from drata controls failingdrata_monitor: failed automated checks from drata monitors failingdrata_connection: disconnected, failed, or never-connected integrationsdrata_personnel: personnel/device compliance issuesdrata_evidence: stale or expiring evidence from drata evidence expiringinconclusive means a drata-cli workflow failed or permissions were insufficient.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-02 | fail→fail | 8,528 | 3,186 | -63% | 1 | 1 | 0% | 1,472 | 713 | -52% | 0 | 0 | — |
case-01 | fail→fail | 8,617 | 24,737 | +187% | 1 | 1 | 0% | 1,244 | 3,130 | +152% | 0 | 0 | — |
case-03 | fail→fail | 3,384 | 2,791 | -18% | 1 | 1 | 0% | 466 | 674 | +45% | 0 | 0 | — |
case-04 | pass→pass | 14,008 | 5,978 | -57% | 1 | 1 | 0% | 2,142 | 1,311 | -39% | 0 | 0 | — |
case-05 | pass→pass | 11,788 | 9,227 | -22% | 1 | 1 | 0% | 1,790 | 1,787 | -0% | 0 | 0 | — |
case-06 | fail→pass | 14,639 | 1,936 | -87% | 1 | 1 | 0% | 2,400 | 602 | -75% | 0 | 0 | — |
case-07 | pass→pass | 8,901 | 3,304 | -63% | 1 | 1 | 0% | 1,489 | 900 | -40% | 0 | 0 | — |
case-08 | fail→pass | 18,049 | 1,810 | -90% | 1 | 1 | 0% | 2,687 | 654 | -76% | 0 | 0 | — |
case-09 | fail→pass | 17,880 | 2,043 | -89% | 1 | 1 | 0% | 2,210 | 703 | -68% | 0 | 0 | — |
case-10 | fail→pass | 9,548 | 2,178 | -77% | 1 | 1 | 0% | 1,493 | 745 | -50% | 0 | 0 | — |
case-11 | fail→pass | 12,817 | 4,596 | -64% | 1 | 1 | 0% | 2,088 | 1,129 | -46% | 0 | 0 | — |
case-12 | fail→pass | 12,174 | 1,427 | -88% | 1 | 1 | 0% | 2,069 | 597 | -71% | 0 | 0 | — |
case-13 | fail→pass | 11,909 | 1,638 | -86% | 1 | 1 | 0% | 1,879 | 576 | -69% | 0 | 0 | — |
case-14 | pass→pass | 12,510 | 5,101 | -59% | 1 | 1 | 0% | 1,858 | 1,119 | -40% | 0 | 0 | — |
case-15 | pass→pass | 15,186 | 7,450 | -51% | 1 | 1 | 0% | 2,202 | 1,506 | -32% | 0 | 0 | — |
case-16 | pass→pass | 15,467 | 8,076 | -48% | 1 | 1 | 0% | 2,187 | 1,514 | -31% | 0 | 0 | — |
case-17 | pass→pass | 9,375 | 9,552 | +2% | 1 | 1 | 0% | 1,298 | 1,682 | +30% | 0 | 0 | — |
case-18 | pass→pass | 16,151 | 9,963 | -38% | 1 | 1 | 0% | 2,232 | 1,737 | -22% | 0 | 0 | — |
case-19 | pass→pass | 14,818 | 6,196 | -58% | 1 | 1 | 0% | 2,638 | 1,387 | -47% | 0 | 0 | — |
case-20 | fail→pass | 15,301 | 1,893 | -88% | 1 | 1 | 0% | 2,468 | 654 | -74% | 0 | 0 | — |
case-21 | pass→pass | 16,783 | 9,933 | -41% | 1 | 1 | 0% | 2,401 | 1,839 | -23% | 0 | 0 | — |
case-22 | pass→pass | 9,511 | 6,441 | -32% | 1 | 1 | 0% | 1,622 | 1,488 | -8% | 0 | 0 | — |
case-23 | pass→pass | 21,055 | 10,586 | -50% | 1 | 1 | 0% | 3,912 | 2,268 | -42% | 0 | 0 | — |
case-24 | pass→pass | 15,109 | 14,352 | -5% | 1 | 1 | 0% | 2,383 | 2,552 | +7% | 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. 24 cases were attempted. The headline lift of +33 percentage points is the difference between those two pass rates over the 24 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.