Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Generate operational runbooks from a service name — deployment, incident response, maintenance, and rollback workflows. Templated structure customizable per environment. Use when documenting on-call procedures for a new service, standardizing incident response across teams, or producing runbooks before launching to production.
.claude/skills/alirezarezvani-runbook-generator/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | -13% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -14% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -41% | 0% |
| case-13 | ✗→✓ | ▲ Improved | 24% | 0% |
| case-16 | ✗→✓ | ▲ Improved | -22% | 0% |
Tier: POWERFUL Category: Engineering Domain: DevOps / Site Reliability Engineering
Generate operational runbooks quickly from a service name, then customize for deployment, incident response, maintenance, and rollback workflows.
bash# Print runbook to stdout python3 scripts/runbook_generator.py payments-api # Write runbook file python3 scripts/runbook_generator.py payments-api --owner platform --output docs/runbooks/payments-api.md
scripts/runbook_generator.py.references/runbook-templates.md| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 18,780 | 18,658 | -1% | 1 | 1 | 0% | 3,124 | 3,907 | +25% | 0 | 0 | — |
case-02 | fail→fail | 17,574 | 18,413 | +5% | 1 | 1 | 0% | 3,162 | 3,689 | +17% | 0 | 0 | — |
case-03 | fail→fail | 16,487 | 15,833 | -4% | 1 | 1 | 0% | 3,165 | 3,485 | +10% | 0 | 0 | — |
case-04 | fail→pass | 5,053 | 2,251 | -55% | 1 | 1 | 0% | 1,005 | 871 | -13% | 0 | 0 | — |
case-05 | fail→pass | 5,125 | 2,075 | -60% | 1 | 1 | 0% | 902 | 775 | -14% | 0 | 0 | — |
case-06 | pass→fail | 12,145 | 7,967 | -34% | 1 | 1 | 0% | 2,133 | 1,849 | -13% | 0 | 0 | — |
case-07 | pass→pass | 8,348 | 2,974 | -64% | 1 | 1 | 0% | 1,440 | 900 | -38% | 0 | 0 | — |
case-08 | fail→pass | 7,396 | 2,718 | -63% | 1 | 1 | 0% | 1,392 | 827 | -41% | 0 | 0 | — |
case-09 | pass→pass | 13,647 | 12,449 | -9% | 1 | 1 | 0% | 2,373 | 2,598 | +9% | 0 | 0 | — |
case-10 | pass→pass | 11,988 | 10,757 | -10% | 1 | 1 | 0% | 2,017 | 2,375 | +18% | 0 | 0 | — |
case-11 | pass→pass | 10,467 | 6,497 | -38% | 1 | 1 | 0% | 1,819 | 1,556 | -14% | 0 | 0 | — |
case-12 | pass→pass | 13,745 | 15,292 | +11% | 1 | 1 | 0% | 2,271 | 2,983 | +31% | 0 | 0 | — |
case-13 | fail→pass | 12,566 | 13,323 | +6% | 1 | 1 | 0% | 2,096 | 2,595 | +24% | 0 | 0 | — |
case-14 | pass→pass | 7,814 | 6,970 | -11% | 1 | 1 | 0% | 1,364 | 1,552 | +14% | 0 | 0 | — |
case-15 | pass→pass | 4,980 | 2,125 | -57% | 1 | 1 | 0% | 859 | 775 | -10% | 0 | 0 | — |
case-16 | fail→pass | 7,296 | 3,117 | -57% | 1 | 1 | 0% | 1,263 | 981 | -22% | 0 | 0 | — |
case-17 | pass→pass | 9,181 | 7,572 | -18% | 1 | 1 | 0% | 1,445 | 1,594 | +10% | 0 | 0 | — |
case-18 | pass→pass | 9,063 | 4,087 | -55% | 1 | 1 | 0% | 1,473 | 1,090 | -26% | 0 | 0 | — |
case-19 | pass→pass | 10,070 | 6,961 | -31% | 1 | 1 | 0% | 1,580 | 1,572 | -1% | 0 | 0 | — |
case-20 | pass→pass | 11,042 | 10,991 | -0% | 1 | 1 | 0% | 2,263 | 2,787 | +23% | 0 | 0 | — |
case-21 | pass→pass | 8,054 | 8,916 | +11% | 1 | 1 | 0% | 1,723 | 2,350 | +36% | 0 | 0 | — |
case-22 | pass→pass | 4,031 | 4,916 | +22% | 1 | 1 | 0% | 911 | 1,463 | +61% | 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 +18 percentage points is the difference between those two pass rates over the 22 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.