Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →You are an API mocking expert specializing in realistic mock services for development, testing, and demos. Design mocks that simulate real API behavior and enable parallel development.
.claude/skills/dokhacgiakhoa-api-testing-observability-api-mock/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | 4% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 42% | 0% |
| case-16 | ✗→✓ | ▲ Improved | 74% | 0% |
| case-20 | ✗→✓ | ▲ Improved | 26% | 0% |
| case-07 | ✓→✗ | ▼ Worse | 87% | 0% |
You are an API mocking expert specializing in creating realistic mock services for development, testing, and demonstration purposes. Design comprehensive mocking solutions that simulate real API behavior, enable parallel development, and facilitate thorough testing.
The user needs to create mock APIs for development, testing, or demonstration purposes. Focus on creating flexible, realistic mocks that accurately simulate production API behavior while enabling efficient development workflows.
$ARGUMENTS
resources/implementation-playbook.md.resources/implementation-playbook.md for code samples, checklists, and templates.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 31,473 | 29,869 | -5% | 1 | 1 | 0% | 5,876 | 5,947 | +1% | 0 | 0 | — |
case-02 | fail→fail | 33,309 | 28,546 | -14% | 1 | 1 | 0% | 6,514 | 5,560 | -15% | 0 | 0 | — |
case-03 | pass→pass | 20,442 | 21,497 | +5% | 1 | 1 | 0% | 2,519 | 3,479 | +38% | 0 | 0 | — |
case-04 | fail→pass | 16,549 | 16,927 | +2% | 1 | 1 | 0% | 2,305 | 2,400 | +4% | 0 | 0 | — |
case-05 | fail→fail | 20,714 | 25,791 | +25% | 1 | 1 | 0% | 2,700 | 4,325 | +60% | 0 | 0 | — |
case-06 | pass→pass | 15,260 | 29,545 | +94% | 1 | 1 | 0% | 1,761 | 4,029 | +129% | 0 | 0 | — |
case-07 | pass→fail | 15,247 | 24,521 | +61% | 1 | 1 | 0% | 2,226 | 4,153 | +87% | 0 | 0 | — |
case-08 | fail→fail | 40,102 | 30,973 | -23% | 1 | 1 | 0% | 8,232 | 6,426 | -22% | 0 | 0 | — |
case-09 | fail→pass | 17,053 | 20,562 | +21% | 1 | 1 | 0% | 3,115 | 4,432 | +42% | 0 | 0 | — |
case-10 | pass→pass | 14,273 | 14,811 | +4% | 1 | 1 | 0% | 2,616 | 3,201 | +22% | 0 | 0 | — |
case-11 | pass→pass | 19,873 | 19,726 | -1% | 1 | 1 | 0% | 2,596 | 2,999 | +16% | 0 | 0 | — |
case-12 | pass→pass | 15,333 | 18,785 | +23% | 1 | 1 | 0% | 2,625 | 2,710 | +3% | 0 | 0 | — |
case-13 | pass→pass | 22,220 | 21,951 | -1% | 1 | 1 | 0% | 3,226 | 3,774 | +17% | 0 | 0 | — |
case-14 | fail→fail | 15,243 | 19,797 | +30% | 1 | 1 | 0% | 2,967 | 4,401 | +48% | 0 | 0 | — |
case-15 | pass→pass | 21,410 | 23,519 | +10% | 1 | 1 | 0% | 2,748 | 3,612 | +31% | 0 | 0 | — |
case-16 | fail→pass | 17,108 | 26,572 | +55% | 1 | 1 | 0% | 2,817 | 4,893 | +74% | 0 | 0 | — |
case-17 | pass→pass | 20,892 | 15,695 | -25% | 1 | 1 | 0% | 2,702 | 3,465 | +28% | 0 | 0 | — |
case-18 | pass→pass | 15,542 | 13,413 | -14% | 1 | 1 | 0% | 2,909 | 3,120 | +7% | 0 | 0 | — |
case-19 | pass→pass | 21,207 | 17,873 | -16% | 1 | 1 | 0% | 2,888 | 3,640 | +26% | 0 | 0 | — |
case-20 | fail→pass | 21,515 | 23,260 | +8% | 1 | 1 | 0% | 2,552 | 3,218 | +26% | 0 | 0 | — |
case-21 | pass→pass | 20,769 | 19,247 | -7% | 1 | 1 | 0% | 2,686 | 2,864 | +7% | 0 | 0 | — |
case-22 | fail→fail | 22,819 | 22,236 | -3% | 1 | 1 | 0% | 3,301 | 3,923 | +19% | 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 +14 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.