Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Interpret slack-inspector findings, explain Slack API coverage limits, and turn Slack workspace posture results into control evidence or remediation.
.claude/skills/grcengclub-slack-inspector-expert/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 3% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 60% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 1198% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -61% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -28% | 0% |
Use this skill when reviewing output from slack-inspector or planning remediation for Slack workspace security.
slack-inspector writes Finding schema v1 documents to:
text~/.cache/claude-grc/findings/slack-inspector/<run_id>.json
Each run emits a workspace-level slack_workspace resource with SCF evaluations for:
IAC-02: MFA / 2FA evidenceIAC-04: SSO and session-duration evidenceDCH-07: message and file retention visibilityTPM-03: app approval workflow visibilityDLP-01: DLP or discovery visibilityGOV-03: workspace metadata inventorySlack exposes many workspace governance controls only through Enterprise Grid or admin-scoped APIs. Treat inconclusive as a coverage signal, not as a pass:
missing_scope, not_allowed_token_type, or not_authed: request an admin-approved token with the listed scope.feature_not_enabled or enterprise_required: verify manually in Slack Enterprise Admin settings or document plan limitations.channel_not_found on retention checks: configure a representative channel for retention policy sampling.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-21 | pass→pass | 19,031 | 14,138 | -26% | 1 | 1 | 0% | 2,720 | 2,320 | -15% | 0 | 0 | — |
case-01 | fail→pass | 45,723 | 25,732 | -44% | 1 | 1 | 0% | 4,731 | 4,869 | +3% | 0 | 0 | — |
case-02 | fail→pass | 18,742 | 25,127 | +34% | 1 | 1 | 0% | 2,895 | 4,624 | +60% | 0 | 0 | — |
case-03 | fail→pass | 5,483 | 15,573 | +184% | 1 | 1 | 0% | 244 | 3,168 | +1198% | 0 | 0 | — |
case-04 | pass→pass | 9,267 | 7,009 | -24% | 1 | 1 | 0% | 1,850 | 1,714 | -7% | 0 | 0 | — |
case-05 | pass→pass | 12,015 | 9,125 | -24% | 1 | 1 | 0% | 2,427 | 2,189 | -10% | 0 | 0 | — |
case-06 | pass→pass | 12,782 | 10,694 | -16% | 1 | 1 | 0% | 2,275 | 2,357 | +4% | 0 | 0 | — |
case-07 | fail→pass | 12,262 | 1,937 | -84% | 1 | 1 | 0% | 1,957 | 770 | -61% | 0 | 0 | — |
case-08 | fail→pass | 22,811 | 4,289 | -81% | 1 | 1 | 0% | 1,531 | 1,101 | -28% | 0 | 0 | — |
case-09 | pass→pass | 10,367 | 1,718 | -83% | 1 | 1 | 0% | 1,679 | 669 | -60% | 0 | 0 | — |
case-10 | fail→pass | 14,103 | 5,305 | -62% | 1 | 1 | 0% | 2,205 | 1,223 | -45% | 0 | 0 | — |
case-11 | fail→pass | 11,491 | 2,088 | -82% | 1 | 1 | 0% | 1,778 | 711 | -60% | 0 | 0 | — |
case-12 | pass→pass | 12,787 | 6,032 | -53% | 1 | 1 | 0% | 1,995 | 1,364 | -32% | 0 | 0 | — |
case-13 | fail→pass | 11,160 | 5,985 | -46% | 1 | 1 | 0% | 1,809 | 1,335 | -26% | 0 | 0 | — |
case-14 | pass→pass | 12,273 | 6,838 | -44% | 1 | 1 | 0% | 1,946 | 1,399 | -28% | 0 | 0 | — |
case-15 | fail→pass | 10,921 | 6,541 | -40% | 1 | 1 | 0% | 1,635 | 1,398 | -14% | 0 | 0 | — |
case-16 | pass→pass | 15,598 | 5,941 | -62% | 1 | 1 | 0% | 2,188 | 1,286 | -41% | 0 | 0 | — |
case-17 | pass→pass | 13,675 | 8,559 | -37% | 1 | 1 | 0% | 1,924 | 1,648 | -14% | 0 | 0 | — |
case-18 | pass→pass | 12,817 | 8,724 | -32% | 1 | 1 | 0% | 1,818 | 1,715 | -6% | 0 | 0 | — |
case-19 | pass→pass | 16,104 | 10,243 | -36% | 1 | 1 | 0% | 2,473 | 1,955 | -21% | 0 | 0 | — |
case-20 | fail→fail | 13,096 | 6,853 | -48% | 1 | 1 | 0% | 2,024 | 1,415 | -30% | 0 | 0 | — |
case-22 | pass→pass | 15,663 | 9,992 | -36% | 1 | 1 | 0% | 2,299 | 1,814 | -21% | 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 20 counted toward the lift figure. The other 2 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 +41 percentage points is the difference between those two pass rates over the 20 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.