Install any skill in seconds. Free to start, no credit card required.
Get Started Free →View all tracked vulnerabilities and their current status
.claude/skills/bilal140202-dashboard/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | -5% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 19% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 34% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 17% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 8% | 0% |
This skill reads .vulnetix/memory.yaml and displays a comprehensive vulnerability status report. It is read-only and does not modify any files.
.vulnetix/memory.yaml exists in the repo root/vulnetix:vuln <package> or /vulnetix:exploits-search to start tracking." and stop..vulnetix/memory.yamlFrom the vulnerabilities: section, categorize each entry:
Open (unresolved):
status: affected -- "Vulnerable"status: under_investigation -- "Investigating"Resolved:
status: fixed -- "Fixed"status: not_affected -- "Not affected"decision.choice: risk-accepted -- "Risk accepted"decision.choice: deferred -- "Deferred"From the manifests: section, collect manifest tracking info.
Vulnetix Security Dashboard
============================
Open: <N> (<X> vulnerable, <Y> investigating)
Resolved: <N> (<X> fixed, <Y> not affected, <Z> risk-accepted, <W> deferred)
Manifests tracked: <N> (last scan: <timestamp>)If there are zero vulnerabilities and zero manifests, display: "Clean slate -- no vulnerabilities tracked yet."
If there are open vulnerabilities, display them sorted by CWSS priority (P1 first), then by severity:
Open Vulnerabilities
--------------------
| ID | Package | Severity | Status | Priority | Decision |
|----|---------|----------|--------|----------|----------|
| CVE-2021-44228 | log4j-core | critical | Vulnerable | P1 (87.5) | investigating |
| GHSA-xxxx-yyyy | express | high | Investigating | P2 (62.0) | investigating |For each column:
package fieldseverity fieldcwss.priority and cwss.score if available, otherwise "--"decision.choice if available, otherwise "--"If there are resolved vulnerabilities, display them:
Resolved Vulnerabilities
------------------------
| ID | Package | Severity | Resolution | Decision | Date |
|----|---------|----------|------------|----------|------|
| CVE-2023-1234 | lodash | high | Fixed | fix-applied | 2024-01-15 |For the Date column, use the most recent history entry timestamp, or discovery.date as fallback.
If manifests are tracked, display:
Tracked Manifests
-----------------
| Manifest | Ecosystem | Last Scanned | Vulns Found |
|----------|-----------|--------------|-------------|
| package.json | npm | 2024-01-15T10:30:00Z | 3 |
| go.mod | go | 2024-01-15T10:31:00Z | 0 |For each open vulnerability (up to 5), suggest a next action based on its state:
threat_model or cwss: "/vulnetix:exploits <id>" -- get exploit analysis and priority scoringcwss but no fix applied: "/vulnetix:fix <id>" -- get fix intelligence"/vulnetix:remediation <id>" -- get a full remediation planIf there are more than 5 open vulns, add: "Use /vulnetix:exploits-search to find exploited vulnerabilities across your ecosystem."
Always end with: "Use /vulnetix:vuln <id> for detailed info on any vulnerability."
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 19,711 | 6,673 | -66% | 1 | 1 | 0% | 4,141 | 1,189 | -71% | 0 | 0 | — |
case-02 | fail→fail | 15,065 | 3,599 | -76% | 1 | 1 | 0% | 3,193 | 1,173 | -63% | 0 | 0 | — |
case-03 | fail→fail | 10,719 | 17,917 | +67% | 1 | 1 | 0% | 2,189 | 1,393 | -36% | 0 | 0 | — |
case-04 | fail→pass | 7,683 | 2,577 | -66% | 1 | 1 | 0% | 1,486 | 1,418 | -5% | 0 | 0 | — |
case-05 | pass→pass | 7,988 | 2,713 | -66% | 1 | 1 | 0% | 1,364 | 1,388 | +2% | 0 | 0 | — |
case-06 | fail→pass | 6,380 | 1,732 | -73% | 1 | 1 | 0% | 1,175 | 1,393 | +19% | 0 | 0 | — |
case-07 | fail→pass | 7,370 | 3,758 | -49% | 1 | 1 | 0% | 1,351 | 1,810 | +34% | 0 | 0 | — |
case-08 | fail→pass | 5,881 | 2,085 | -65% | 1 | 1 | 0% | 1,289 | 1,506 | +17% | 0 | 0 | — |
case-09 | pass→pass | 5,287 | 2,525 | -52% | 1 | 1 | 0% | 1,125 | 1,496 | +33% | 0 | 0 | — |
case-10 | fail→pass | 14,725 | 1,576 | -89% | 1 | 1 | 0% | 1,213 | 1,311 | +8% | 0 | 0 | — |
case-11 | fail→pass | 7,397 | 1,633 | -78% | 1 | 1 | 0% | 1,241 | 1,293 | +4% | 0 | 0 | — |
case-12 | fail→pass | 6,307 | 2,090 | -67% | 1 | 1 | 0% | 1,048 | 1,270 | +21% | 0 | 0 | — |
case-13 | fail→pass | 10,115 | 3,152 | -69% | 1 | 1 | 0% | 2,089 | 1,593 | -24% | 0 | 0 | — |
case-14 | fail→pass | 8,969 | 2,851 | -68% | 1 | 1 | 0% | 1,783 | 1,524 | -15% | 0 | 0 | — |
case-15 | fail→pass | 8,806 | 2,284 | -74% | 1 | 1 | 0% | 1,505 | 1,406 | -7% | 0 | 0 | — |
case-16 | fail→pass | 10,347 | 3,568 | -66% | 1 | 1 | 0% | 1,665 | 1,643 | -1% | 0 | 0 | — |
case-17 | fail→pass | 8,615 | 1,923 | -78% | 1 | 1 | 0% | 1,217 | 1,323 | +9% | 0 | 0 | — |
case-18 | fail→pass | 8,625 | 3,701 | -57% | 1 | 1 | 0% | 1,266 | 1,147 | -9% | 0 | 0 | — |
case-19 | fail→pass | 9,787 | 1,797 | -82% | 1 | 1 | 0% | 1,463 | 1,317 | -10% | 0 | 0 | — |
case-20 | pass→pass | 8,247 | 3,355 | -59% | 1 | 1 | 0% | 1,370 | 1,401 | +2% | 0 | 0 | — |
case-21 | fail→fail | 6,288 | 7,042 | +12% | 1 | 1 | 0% | 998 | 1,404 | +41% | 0 | 0 | — |
case-22 | fail→fail | 5,108 | 10,246 | +101% | 1 | 1 | 0% | 206 | 1,596 | +675% | 0 | 0 | — |
case-23 | pass→pass | 5,140 | 5,573 | +8% | 1 | 1 | 0% | 365 | 1,326 | +263% | 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. 23 cases were attempted, and 17 counted toward the lift figure. The other 6 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 +61 percentage points is the difference between those two pass rates over the 17 comparable cases. 2 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.