Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Generate technical implementation plans from feature specifications. Use after creating a spec to define architecture, tech stack, and implementation phases. Creates plan.md with detailed technical design.
.claude/skills/foryourhealth111-pixel-speckit-plan/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-05 | ✗→✓ | ▲ Improved | -30% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -19% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -36% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -53% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -27% | 0% |
text$ARGUMENTS
You MUST consider the user input before proceeding (if not empty).
.specify/scripts/powershell/setup-plan.ps1 -Json from repo root and parse JSON for FEATURE_SPEC, IMPL_PLAN, SPECS_DIR, BRANCH. For single quotes in args like "I'm Groot", use escape syntax: e.g 'I'\''m Groot' (or double-quote if possible: "I'm Groot")..specify/memory/constitution.md. Load IMPL_PLAN template (already copied).text For each unknown in Technical Context: Task: "Research {unknown} for {feature context}" For each technology choice: Task: "Find best practices for {tech} in {domain}"
research.md using format:Output: research.md with all NEEDS CLARIFICATION resolved
Prerequisites: research.md complete
data-model.md:/contracts/:.specify/scripts/powershell/update-agent-context.ps1 -AgentType codexOutput: data-model.md, /contracts/, quickstart.md, agent-specific file
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 27,008 | 5,108 | -81% | 1 | 1 | 0% | 5,159 | 981 | -81% | 0 | 0 | — |
case-02 | fail→fail | 28,920 | 4,643 | -84% | 1 | 1 | 0% | 6,186 | 963 | -84% | 0 | 0 | — |
case-03 | fail→fail | 33,586 | 5,276 | -84% | 1 | 1 | 0% | 6,181 | 1,011 | -84% | 0 | 0 | — |
case-04 | pass→pass | 6,401 | 3,716 | -42% | 1 | 1 | 0% | 1,179 | 1,390 | +18% | 0 | 0 | — |
case-05 | fail→pass | 11,607 | 4,830 | -58% | 1 | 1 | 0% | 2,031 | 1,427 | -30% | 0 | 0 | — |
case-06 | fail→pass | 12,429 | 5,366 | -57% | 1 | 1 | 0% | 2,028 | 1,642 | -19% | 0 | 0 | — |
case-07 | fail→pass | 10,779 | 2,243 | -79% | 1 | 1 | 0% | 1,748 | 1,113 | -36% | 0 | 0 | — |
case-08 | fail→pass | 14,804 | 6,557 | -56% | 1 | 1 | 0% | 2,549 | 1,196 | -53% | 0 | 0 | — |
case-09 | fail→pass | 10,886 | 2,846 | -74% | 1 | 1 | 0% | 1,721 | 1,256 | -27% | 0 | 0 | — |
case-10 | fail→pass | 6,986 | 2,631 | -62% | 1 | 1 | 0% | 1,169 | 1,186 | +1% | 0 | 0 | — |
case-11 | pass→pass | 6,410 | 3,257 | -49% | 1 | 1 | 0% | 1,053 | 1,194 | +13% | 0 | 0 | — |
case-12 | pass→pass | 8,191 | 3,320 | -59% | 1 | 1 | 0% | 1,201 | 1,295 | +8% | 0 | 0 | — |
case-13 | fail→fail | 8,883 | 2,758 | -69% | 1 | 1 | 0% | 1,490 | 1,174 | -21% | 0 | 0 | — |
case-14 | fail→pass | 5,686 | 1,422 | -75% | 1 | 1 | 0% | 995 | 954 | -4% | 0 | 0 | — |
case-15 | pass→pass | 12,928 | 2,877 | -78% | 1 | 1 | 0% | 2,072 | 1,197 | -42% | 0 | 0 | — |
case-16 | pass→pass | 11,493 | 3,367 | -71% | 1 | 1 | 0% | 1,740 | 1,364 | -22% | 0 | 0 | — |
case-17 | pass→pass | 12,706 | 3,330 | -74% | 1 | 1 | 0% | 2,161 | 1,207 | -44% | 0 | 0 | — |
case-18 | pass→pass | 11,646 | 3,749 | -68% | 1 | 1 | 0% | 1,839 | 1,394 | -24% | 0 | 0 | — |
case-19 | fail→pass | 9,197 | 2,805 | -70% | 1 | 1 | 0% | 1,401 | 1,295 | -8% | 0 | 0 | — |
case-20 | pass→fail | 2,510 | 24,795 | +888% | 1 | 1 | 0% | 414 | 4,778 | +1054% | 0 | 0 | — |
case-21 | pass→fail | 21,394 | 6,956 | -67% | 1 | 1 | 0% | 4,640 | 1,101 | -76% | 0 | 0 | — |
case-22 | pass→fail | 32,435 | 6,844 | -79% | 1 | 1 | 0% | 2,794 | 1,110 | -60% | 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 +23 percentage points is the difference between those two pass rates over the 17 comparable cases. 3 cases got worse with the skill loaded, and they are 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.