Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Test-first development route for TDD, writing failing tests first, RED -> GREEN -> REFACTOR, and behavior-changing feature/bug/refactor work. Do not use for already-failing test root-cause debugging, test-report packaging, or final completion evidence.
.claude/skills/foryourhealth111-pixel-tdd-guide/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-05 | ✗→✓ | ▲ Improved | -51% | 0% |
| case-08 | ✓→✗ | ▼ Worse | -61% | 0% |
| case-02 | ✓→✓ | ＝ Same ✓ | 81% | 0% |
| case-03 | ✓→✓ | ＝ Same ✓ | 98% | 0% |
| case-04 | ✓→✓ | ＝ Same ✓ | 40% | 0% |
Use this skill for all feature work, bug fixes, and refactors that change behavior.
Use this skill when the user wants to build or change behavior through tests:
Do not use it when tests are already failing and the user wants root cause; route that to systematic-debugging. Do not use it when the user only wants a test/coverage report; route that to generating-test-reports.
No production code before a failing test.
everything-claude-code:tdd-guide.test-driven-development.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 20,090 | 11,859 | -41% | 1 | 1 | 0% | 4,267 | 2,834 | -34% | 0 | 0 | — |
case-02 | pass→pass | 2,796 | 2,567 | -8% | 1 | 1 | 0% | 383 | 695 | +81% | 0 | 0 | — |
case-03 | pass→pass | 2,124 | 2,627 | +24% | 1 | 1 | 0% | 352 | 697 | +98% | 0 | 0 | — |
case-04 | pass→pass | 4,209 | 3,506 | -17% | 1 | 1 | 0% | 670 | 941 | +40% | 0 | 0 | — |
case-05 | fail→pass | 10,637 | 4,718 | -56% | 1 | 1 | 0% | 2,246 | 1,104 | -51% | 0 | 0 | — |
case-06 | pass→pass | 8,030 | 3,887 | -52% | 1 | 1 | 0% | 1,397 | 902 | -35% | 0 | 0 | — |
case-07 | pass→pass | 8,437 | 4,428 | -48% | 1 | 1 | 0% | 1,448 | 1,142 | -21% | 0 | 0 | — |
case-08 | pass→fail | 10,680 | 2,801 | -74% | 1 | 1 | 0% | 2,021 | 798 | -61% | 0 | 0 | — |
case-09 | pass→pass | 11,092 | 7,473 | -33% | 1 | 1 | 0% | 1,847 | 1,453 | -21% | 0 | 0 | — |
case-10 | fail→fail | 12,724 | 10,037 | -21% | 1 | 1 | 0% | 2,162 | 1,987 | -8% | 0 | 0 | — |
case-11 | pass→pass | 16,364 | 11,139 | -32% | 1 | 1 | 0% | 2,510 | 2,056 | -18% | 0 | 0 | — |
case-12 | pass→pass | 15,066 | 11,554 | -23% | 1 | 1 | 0% | 2,483 | 2,250 | -9% | 0 | 0 | — |
case-13 | pass→pass | 8,109 | 6,153 | -24% | 1 | 1 | 0% | 1,555 | 1,414 | -9% | 0 | 0 | — |
case-14 | pass→pass | 9,847 | 5,507 | -44% | 1 | 1 | 0% | 1,582 | 1,289 | -19% | 0 | 0 | — |
case-15 | pass→pass | 13,845 | 9,189 | -34% | 1 | 1 | 0% | 2,323 | 1,998 | -14% | 0 | 0 | — |
case-16 | pass→pass | 9,846 | 7,339 | -25% | 1 | 1 | 0% | 2,041 | 1,861 | -9% | 0 | 0 | — |
case-17 | pass→pass | 13,461 | 11,360 | -16% | 1 | 1 | 0% | 2,585 | 2,336 | -10% | 0 | 0 | — |
case-18 | pass→pass | 15,376 | 10,453 | -32% | 1 | 1 | 0% | 2,710 | 2,199 | -19% | 0 | 0 | — |
case-19 | pass→pass | 9,175 | 5,899 | -36% | 1 | 1 | 0% | 1,498 | 1,364 | -9% | 0 | 0 | — |
case-20 | pass→pass | 10,394 | 6,889 | -34% | 1 | 1 | 0% | 1,670 | 1,515 | -9% | 0 | 0 | — |
case-21 | pass→pass | 8,297 | 3,841 | -54% | 1 | 1 | 0% | 1,272 | 1,038 | -18% | 0 | 0 | — |
case-22 | pass→pass | 8,805 | 5,341 | -39% | 1 | 1 | 0% | 1,554 | 1,276 | -18% | 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 0 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.