Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Inngest expert for serverless-first background jobs, event-driven workflows, and durable execution without managing queues or workers. Use when: inngest, serverless background job, event-driven workflow, step function, durable execution.
.claude/skills/dokhacgiakhoa-inngest/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -14% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -14% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -16% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -3% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -11% | 0% |
You are an Inngest expert who builds reliable background processing without managing infrastructure. You understand that serverless doesn't mean you can't have durable, long-running workflows - it means you don't manage the workers.
You've built AI pipelines that take minutes, onboarding flows that span days, and event-driven systems that process millions of events. You know that the magic of Inngest is in its steps - each one a checkpoint that survives failures.
Your core philosophy:
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 18,181 | 34,452 | +89% | 1 | 1 | 0% | 2,475 | 2,127 | -14% | 0 | 0 | — |
case-02 | pass→pass | 13,026 | 8,530 | -35% | 1 | 1 | 0% | 2,038 | 1,919 | -6% | 0 | 0 | — |
case-03 | pass→pass | 11,801 | 9,751 | -17% | 1 | 1 | 0% | 1,845 | 1,692 | -8% | 0 | 0 | — |
case-04 | pass→pass | 13,222 | 20,195 | +53% | 1 | 1 | 0% | 2,120 | 1,989 | -6% | 0 | 0 | — |
case-05 | fail→pass | 8,072 | 5,139 | -36% | 1 | 1 | 0% | 1,380 | 1,181 | -14% | 0 | 0 | — |
case-06 | fail→pass | 21,021 | 13,016 | -38% | 1 | 1 | 0% | 3,104 | 2,597 | -16% | 0 | 0 | — |
case-07 | fail→pass | 13,949 | 9,128 | -35% | 1 | 1 | 0% | 1,937 | 1,882 | -3% | 0 | 0 | — |
case-08 | fail→pass | 13,592 | 7,957 | -41% | 1 | 1 | 0% | 2,104 | 1,870 | -11% | 0 | 0 | — |
case-09 | fail→pass | 18,114 | 16,984 | -6% | 1 | 1 | 0% | 2,960 | 2,694 | -9% | 0 | 0 | — |
case-10 | fail→pass | 24,520 | 19,742 | -19% | 1 | 1 | 0% | 4,205 | 1,076 | -74% | 0 | 0 | — |
case-11 | pass→pass | 9,419 | 6,034 | -36% | 1 | 1 | 0% | 1,438 | 1,169 | -19% | 0 | 0 | — |
case-12 | fail→pass | 14,107 | 9,632 | -32% | 1 | 1 | 0% | 2,489 | 1,721 | -31% | 0 | 0 | — |
case-13 | pass→pass | 10,142 | 6,623 | -35% | 1 | 1 | 0% | 2,155 | 1,559 | -28% | 0 | 0 | — |
case-14 | fail→pass | 15,154 | 13,754 | -9% | 1 | 1 | 0% | 2,111 | 2,385 | +13% | 0 | 0 | — |
case-15 | fail→pass | 14,244 | 13,920 | -2% | 1 | 1 | 0% | 2,106 | 2,421 | +15% | 0 | 0 | — |
case-16 | pass→pass | 15,424 | 10,548 | -32% | 1 | 1 | 0% | 3,188 | 2,642 | -17% | 0 | 0 | — |
case-17 | fail→pass | 12,181 | 8,079 | -34% | 1 | 1 | 0% | 2,107 | 1,816 | -14% | 0 | 0 | — |
case-18 | fail→pass | 10,106 | 4,666 | -54% | 1 | 1 | 0% | 1,583 | 1,053 | -33% | 0 | 0 | — |
case-19 | pass→pass | 16,705 | 9,485 | -43% | 1 | 1 | 0% | 2,904 | 1,952 | -33% | 0 | 0 | — |
case-20 | pass→pass | 11,595 | 10,653 | -8% | 1 | 1 | 0% | 2,167 | 1,837 | -15% | 0 | 0 | — |
case-21 | pass→pass | 15,661 | 11,641 | -26% | 1 | 1 | 0% | 3,069 | 2,437 | -21% | 0 | 0 | — |
case-22 | pass→fail | 13,818 | 8,010 | -42% | 1 | 1 | 0% | 2,281 | 1,967 | -14% | 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 +50 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.