Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →You are an SLO (Service Level Objective) expert specializing in implementing reliability standards and error budget-based practices. Design SLO frameworks, define SLIs, and build monitoring that balances reliability with delivery velocity.
.claude/skills/dokhacgiakhoa-observability-monitoring-slo-implement/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-23 | ✗→✓ | ▲ Improved | 13% | 0% |
| case-17 | ✓→✗ | ▼ Worse | 13% | 0% |
| case-03 | ✓→✓ | ＝ Same ✓ | 0% | 0% |
| case-04 | ✓→✓ | ＝ Same ✓ | 14% | 0% |
| case-05 | ✓→✓ | ＝ Same ✓ | -16% | 0% |
You are an SLO (Service Level Objective) expert specializing in implementing reliability standards and error budget-based engineering practices. Design comprehensive SLO frameworks, establish meaningful SLIs, and create monitoring systems that balance reliability with feature velocity.
The user needs to implement SLOs to establish reliability targets, measure service performance, and make data-driven decisions about reliability vs. feature development. Focus on practical SLO implementation that aligns with business objectives.
$ARGUMENTS
resources/implementation-playbook.md.resources/implementation-playbook.md for detailed patterns and examples.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 35,612 | 29,097 | -18% | 1 | 1 | 0% | 6,922 | 3,947 | -43% | 0 | 0 | — |
case-02 | fail→fail | 39,887 | 27,942 | -30% | 1 | 1 | 0% | 8,258 | 5,851 | -29% | 0 | 0 | — |
case-03 | pass→pass | 14,042 | 12,943 | -8% | 1 | 1 | 0% | 2,710 | 2,700 | -0% | 0 | 0 | — |
case-04 | pass→pass | 14,044 | 13,979 | -0% | 1 | 1 | 0% | 2,434 | 2,781 | +14% | 0 | 0 | — |
case-05 | pass→pass | 12,792 | 8,725 | -32% | 1 | 1 | 0% | 2,616 | 2,191 | -16% | 0 | 0 | — |
case-06 | fail→fail | 14,126 | 17,608 | +25% | 1 | 1 | 0% | 2,656 | 3,442 | +30% | 0 | 0 | — |
case-07 | fail→fail | 21,113 | 22,253 | +5% | 1 | 1 | 0% | 3,903 | 4,480 | +15% | 0 | 0 | — |
case-08 | pass→pass | 10,791 | 12,343 | +14% | 1 | 1 | 0% | 1,630 | 2,470 | +52% | 0 | 0 | — |
case-09 | pass→pass | 10,378 | 12,025 | +16% | 1 | 1 | 0% | 1,801 | 2,175 | +21% | 0 | 0 | — |
case-10 | fail→fail | 33,846 | 21,592 | -36% | 1 | 1 | 0% | 6,561 | 4,462 | -32% | 0 | 0 | — |
case-11 | pass→pass | 16,134 | 17,002 | +5% | 1 | 1 | 0% | 2,711 | 2,693 | -1% | 0 | 0 | — |
case-12 | pass→pass | 19,568 | 18,211 | -7% | 1 | 1 | 0% | 3,023 | 3,498 | +16% | 0 | 0 | — |
case-13 | fail→fail | 13,455 | 12,943 | -4% | 1 | 1 | 0% | 2,279 | 2,479 | +9% | 0 | 0 | — |
case-14 | fail→fail | 19,106 | 16,130 | -16% | 1 | 1 | 0% | 3,128 | 3,037 | -3% | 0 | 0 | — |
case-15 | fail→fail | 14,656 | 14,852 | +1% | 1 | 1 | 0% | 2,333 | 2,736 | +17% | 0 | 0 | — |
case-16 | fail→fail | 13,461 | 13,923 | +3% | 1 | 1 | 0% | 2,205 | 2,576 | +17% | 0 | 0 | — |
case-17 | pass→fail | 15,438 | 15,368 | -0% | 1 | 1 | 0% | 2,635 | 2,969 | +13% | 0 | 0 | — |
case-18 | pass→pass | 19,217 | 19,117 | -1% | 1 | 1 | 0% | 2,669 | 3,330 | +25% | 0 | 0 | — |
case-19 | fail→fail | 17,710 | 17,698 | -0% | 1 | 1 | 0% | 3,060 | 3,627 | +19% | 0 | 0 | — |
case-20 | pass→pass | 11,702 | 11,546 | -1% | 1 | 1 | 0% | 1,950 | 2,204 | +13% | 0 | 0 | — |
case-21 | fail→fail | 17,850 | 16,376 | -8% | 1 | 1 | 0% | 2,797 | 2,985 | +7% | 0 | 0 | — |
case-22 | fail→fail | 17,249 | 15,606 | -10% | 1 | 1 | 0% | 2,709 | 2,779 | +3% | 0 | 0 | — |
case-23 | fail→pass | 20,273 | 12,624 | -38% | 1 | 1 | 0% | 2,316 | 2,624 | +13% | 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 0 percentage points is the difference between those two pass rates over the 23 comparable cases. 1 case got worse with the skill loaded, and it is 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.