Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Audits code for OWASP Top 10 vulnerabilities
.claude/skills/observal-security-auditor/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✓→✗ | ▼ Worse | -65% | 0% |
| case-08 | ✓→✓ | ＝ Same ✓ | -16% | 0% |
| case-09 | ✓→✓ | ＝ Same ✓ | 11% | 0% |
| case-16 | ✓→✓ | ＝ Same ✓ | -33% | 0% |
| case-17 | ✓→✓ | ＝ Same ✓ | -2% | 0% |
<!-- SPDX-FileCopyrightText: 2026 Lokesh Selvam <lokeshselvam7025@gmail.com> --> <!-- SPDX-License-Identifier: Apache-2.0 -->
When activated, scan provided code for OWASP Top 10 vulnerabilities:
Provide severity ratings (critical/high/medium/low) and remediation steps with corrected code.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-08 | pass→pass | 15,830 | 12,269 | -22% | 1 | 1 | 0% | 2,956 | 2,486 | -16% | 0 | 0 | — |
case-09 | pass→pass | 13,533 | 13,479 | -0% | 1 | 1 | 0% | 2,479 | 2,746 | +11% | 0 | 0 | — |
case-16 | pass→pass | 12,562 | 8,437 | -33% | 1 | 1 | 0% | 2,263 | 1,506 | -33% | 0 | 0 | — |
case-17 | pass→pass | 12,402 | 11,456 | -8% | 1 | 1 | 0% | 2,324 | 2,268 | -2% | 0 | 0 | — |
case-18 | pass→pass | 14,534 | 10,846 | -25% | 1 | 1 | 0% | 2,573 | 2,154 | -16% | 0 | 0 | — |
case-19 | pass→pass | 9,248 | 7,900 | -15% | 1 | 1 | 0% | 1,620 | 1,473 | -9% | 0 | 0 | — |
case-07 | pass→pass | 12,669 | 11,655 | -8% | 1 | 1 | 0% | 2,196 | 2,226 | +1% | 0 | 0 | — |
case-01 | fail→fail | 17,933 | 10,410 | -42% | 1 | 1 | 0% | 1,162 | 956 | -18% | 0 | 0 | — |
case-02 | fail→fail | 24,185 | 8,228 | -66% | 1 | 1 | 0% | 930 | 917 | -1% | 0 | 0 | — |
case-03 | pass→fail | 18,323 | 8,553 | -53% | 1 | 1 | 0% | 2,655 | 916 | -65% | 0 | 0 | — |
case-04 | pass→pass | 23,109 | 10,186 | -56% | 1 | 1 | 0% | 2,272 | 1,860 | -18% | 0 | 0 | — |
case-05 | pass→pass | 11,251 | 8,561 | -24% | 1 | 1 | 0% | 1,915 | 1,623 | -15% | 0 | 0 | — |
case-06 | pass→pass | 23,013 | 12,626 | -45% | 1 | 1 | 0% | 3,488 | 2,441 | -30% | 0 | 0 | — |
case-10 | pass→pass | 13,225 | 8,743 | -34% | 1 | 1 | 0% | 2,339 | 1,623 | -31% | 0 | 0 | — |
case-11 | pass→pass | 12,165 | 11,776 | -3% | 1 | 1 | 0% | 2,193 | 2,244 | +2% | 0 | 0 | — |
case-12 | pass→pass | 12,534 | 11,250 | -10% | 1 | 1 | 0% | 2,339 | 2,061 | -12% | 0 | 0 | — |
case-13 | pass→pass | 22,762 | 17,963 | -21% | 1 | 1 | 0% | 4,181 | 3,572 | -15% | 0 | 0 | — |
case-14 | pass→pass | 12,571 | 10,511 | -16% | 1 | 1 | 0% | 2,325 | 1,825 | -22% | 0 | 0 | — |
case-15 | pass→pass | 18,844 | 13,086 | -31% | 1 | 1 | 0% | 3,390 | 2,544 | -25% | 0 | 0 | — |
case-20 | fail→fail | 2,907 | 2,978 | +2% | 1 | 1 | 0% | 352 | 575 | +63% | 0 | 0 | — |
case-21 | fail→fail | 8,921 | 3,622 | -59% | 1 | 1 | 0% | 1,304 | 681 | -48% | 0 | 0 | — |
case-22 | pass→pass | 15,360 | 13,700 | -11% | 1 | 1 | 0% | 3,135 | 2,793 | -11% | 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 -100 percentage points is the difference between those two pass rates over the 22 comparable cases. 2 cases got worse with the skill loaded, and they are included in that figure.
Other measured skills in the registry, with their headline benchmark lift.