Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Create pytest suites under the no-mocks policy — real data, temp files, subprocess, pytest-httpserver. USE WHEN adding tests, raising coverage, testing new src/ module, or user forbids mocks.
.claude/skills/docxology-template-test-creation/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | -2% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -15% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -36% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 11% | 0% |
| case-10 | ✗→✓ | ▲ Improved | -6% | 0% |
unittest.mock, MagicMock, or patch the unit under test. Use real computation, tmp_path, local HTTP servers, subprocess for CLIs.test_*.py mirroring src modules; parametrize variants; fixed RNG seeds.projects/<n>/tests/ or tests/infra_tests/ following existing layout.requires_latex, requires_ollama)bashuv run pytest path/to/test_file.py -v uv run pytest projects/<project>/tests/ --cov=projects/<project>/src --cov-fail-under=90 uv run pytest tests/infra_tests/path/ --cov=infrastructure --cov-fail-under=60
docs/rules/testing_standards.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 | 12,402 | 5,208 | -58% | 1 | 1 | 0% | 2,698 | 581 | -78% | 0 | 0 | — |
case-02 | fail→fail | 11,470 | 4,931 | -57% | 1 | 1 | 0% | 2,663 | 657 | -75% | 0 | 0 | — |
case-03 | fail→fail | 3,676 | 4,448 | +21% | 1 | 1 | 0% | 187 | 666 | +256% | 0 | 0 | — |
case-04 | fail→pass | 9,526 | 7,737 | -19% | 1 | 1 | 0% | 2,156 | 2,104 | -2% | 0 | 0 | — |
case-05 | fail→fail | 13,055 | 5,632 | -57% | 1 | 1 | 0% | 2,855 | 648 | -77% | 0 | 0 | — |
case-06 | fail→pass | 14,652 | 14,360 | -2% | 1 | 1 | 0% | 3,268 | 2,785 | -15% | 0 | 0 | — |
case-07 | fail→pass | 12,173 | 7,932 | -35% | 1 | 1 | 0% | 2,918 | 1,866 | -36% | 0 | 0 | — |
case-08 | fail→pass | 10,682 | 10,175 | -5% | 1 | 1 | 0% | 2,277 | 2,520 | +11% | 0 | 0 | — |
case-09 | pass→pass | 3,108 | 1,755 | -44% | 1 | 1 | 0% | 463 | 712 | +54% | 0 | 0 | — |
case-10 | fail→pass | 5,544 | 2,971 | -46% | 1 | 1 | 0% | 960 | 901 | -6% | 0 | 0 | — |
case-11 | fail→pass | 3,992 | 1,200 | -70% | 1 | 1 | 0% | 629 | 604 | -4% | 0 | 0 | — |
case-12 | pass→fail | 6,091 | 3,373 | -45% | 1 | 1 | 0% | 1,352 | 793 | -41% | 0 | 0 | — |
case-13 | pass→pass | 13,715 | 11,561 | -16% | 1 | 1 | 0% | 2,940 | 2,285 | -22% | 0 | 0 | — |
case-14 | fail→pass | 15,369 | 24,741 | +61% | 1 | 1 | 0% | 2,819 | 2,587 | -8% | 0 | 0 | — |
case-15 | fail→pass | 12,132 | 5,910 | -51% | 1 | 1 | 0% | 1,932 | 1,508 | -22% | 0 | 0 | — |
case-16 | pass→pass | 3,415 | 2,790 | -18% | 1 | 1 | 0% | 596 | 747 | +25% | 0 | 0 | — |
case-17 | pass→fail | 11,452 | 6,319 | -45% | 1 | 1 | 0% | 2,437 | 775 | -68% | 0 | 0 | — |
case-18 | pass→pass | 5,868 | 2,354 | -60% | 1 | 1 | 0% | 871 | 836 | -4% | 0 | 0 | — |
case-19 | fail→pass | 6,476 | 2,830 | -56% | 1 | 1 | 0% | 1,673 | 788 | -53% | 0 | 0 | — |
case-20 | fail→fail | 32,538 | 26,487 | -19% | 1 | 1 | 0% | 5,932 | 6,101 | +3% | 0 | 0 | — |
case-21 | fail→fail | 18,442 | 23,217 | +26% | 1 | 1 | 0% | 3,960 | 5,612 | +42% | 0 | 0 | — |
case-22 | fail→pass | 8,373 | 2,156 | -74% | 1 | 1 | 0% | 1,665 | 811 | -51% | 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, and 17 counted toward the lift figure. The other 5 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +36 percentage points is the difference between those two pass rates over the 17 comparable cases. 4 cases got worse with the skill loaded, and they are included in that figure.
The publisher has shipped newer versions since this run, so these numbers describe v1, not the version currently listed.
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.