Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Set up Prometheus for comprehensive metric collection, storage, and monitoring of infrastructure and applications. Use when implementing metrics collection, setting up monitoring infrastructure, or configuring alerting systems.
.claude/skills/dokhacgiakhoa-prometheus-configuration/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-05 | ✗→✓ | ▲ Improved | 13% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 44% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 14% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 40% | 0% |
| case-15 | ✗→✓ | ▲ Improved | 39% | 0% |
Complete guide to Prometheus setup, metric collection, scrape configuration, and recording rules.
resources/implementation-playbook.md.Configure Prometheus for comprehensive metric collection, alerting, and monitoring of infrastructure and applications.
┌──────────────┐
│ Applications │ ← Instrumented with client libraries
└──────┬───────┘
│ /metrics endpoint
↓
┌──────────────┐
│ Prometheus │ ← Scrapes metrics periodically
│ Server │
└──────┬───────┘
│
├─→ AlertManager (alerts)
├─→ Grafana (visualization)
└─→ Long-term storage (Thanos/Cortex)| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 10,270 | 34,649 | +237% | 1 | 1 | 0% | 1,975 | 2,361 | +20% | 0 | 0 | — |
case-02 | fail→fail | 14,493 | 9,822 | -32% | 1 | 1 | 0% | 3,129 | 2,431 | -22% | 0 | 0 | — |
case-03 | fail→fail | 12,655 | 10,051 | -21% | 1 | 1 | 0% | 2,432 | 2,391 | -2% | 0 | 0 | — |
case-04 | fail→fail | 6,644 | 7,067 | +6% | 1 | 1 | 0% | 1,213 | 1,738 | +43% | 0 | 0 | — |
case-05 | fail→pass | 7,209 | 6,332 | -12% | 1 | 1 | 0% | 1,351 | 1,526 | +13% | 0 | 0 | — |
case-06 | fail→pass | 5,342 | 5,566 | +4% | 1 | 1 | 0% | 1,046 | 1,504 | +44% | 0 | 0 | — |
case-07 | fail→fail | 9,775 | 10,183 | +4% | 1 | 1 | 0% | 1,776 | 2,501 | +41% | 0 | 0 | — |
case-08 | fail→pass | 14,095 | 14,019 | -1% | 1 | 1 | 0% | 3,004 | 3,432 | +14% | 0 | 0 | — |
case-09 | pass→pass | 9,904 | 8,896 | -10% | 1 | 1 | 0% | 1,911 | 2,223 | +16% | 0 | 0 | — |
case-10 | fail→pass | 5,288 | 5,253 | -1% | 1 | 1 | 0% | 997 | 1,393 | +40% | 0 | 0 | — |
case-11 | pass→pass | 9,106 | 24,449 | +168% | 1 | 1 | 0% | 1,814 | 3,595 | +98% | 0 | 0 | — |
case-12 | pass→pass | 11,547 | 12,220 | +6% | 1 | 1 | 0% | 2,114 | 2,653 | +25% | 0 | 0 | — |
case-13 | pass→pass | 6,858 | 8,229 | +20% | 1 | 1 | 0% | 1,261 | 1,970 | +56% | 0 | 0 | — |
case-14 | pass→pass | 7,867 | 9,136 | +16% | 1 | 1 | 0% | 1,486 | 2,120 | +43% | 0 | 0 | — |
case-15 | fail→pass | 7,150 | 7,363 | +3% | 1 | 1 | 0% | 1,324 | 1,844 | +39% | 0 | 0 | — |
case-16 | fail→pass | 9,929 | 10,225 | +3% | 1 | 1 | 0% | 1,805 | 2,246 | +24% | 0 | 0 | — |
case-17 | fail→pass | 3,201 | 3,380 | +6% | 1 | 1 | 0% | 579 | 1,013 | +75% | 0 | 0 | — |
case-18 | fail→fail | 7,067 | 4,577 | -35% | 1 | 1 | 0% | 997 | 1,157 | +16% | 0 | 0 | — |
case-19 | pass→pass | 7,054 | 6,561 | -7% | 1 | 1 | 0% | 1,025 | 1,290 | +26% | 0 | 0 | — |
case-20 | fail→fail | 15,293 | 16,699 | +9% | 1 | 1 | 0% | 2,633 | 3,863 | +47% | 0 | 0 | — |
case-21 | fail→fail | 10,592 | 15,100 | +43% | 1 | 1 | 0% | 1,973 | 3,261 | +65% | 0 | 0 | — |
case-22 | fail→fail | 18,029 | 22,881 | +27% | 1 | 1 | 0% | 2,944 | 4,143 | +41% | 0 | 0 | — |
case-23 | fail→fail | 8,098 | 10,061 | +24% | 1 | 1 | 0% | 1,462 | 2,000 | +37% | 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. The headline lift of +30 percentage points is the difference between those two pass rates over the 23 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.