Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Add phase to end of current milestone in roadmap
.claude/skills/davepoon-gsd-add-phase/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-07 | ✗→✓ | ▲ Improved | 1% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -23% | 0% |
| case-10 | ✗→✓ | ▲ Improved | -53% | 0% |
| case-11 | ✗→✓ | ▲ Improved | -38% | 0% |
| case-12 | ✗→✓ | ▲ Improved | -42% | 0% |
<objective> Add a new integer phase to the end of the current milestone in the roadmap.
Routes to the add-phase workflow which handles:
</objective>
<execution_context> @${CLAUDE_PLUGIN_ROOT}/workflows/add-phase.md </execution_context>
<context> Arguments: $ARGUMENTS (phase description)
Roadmap and state are resolved in-workflow via init phase-op and targeted tool calls. </context>
<process> Follow the add-phase workflow from @${CLAUDE_PLUGIN_ROOT}/workflows/add-phase.md.
The workflow handles all logic including:
</process>
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 5,447 | 1,944 | -64% | 1 | 1 | 0% | 934 | 531 | -43% | 0 | 0 | — |
case-02 | fail→fail | 3,752 | 5,312 | +42% | 1 | 1 | 0% | 198 | 569 | +187% | 0 | 0 | — |
case-03 | pass→fail | 8,818 | 4,019 | -54% | 1 | 1 | 0% | 1,605 | 496 | -69% | 0 | 0 | — |
case-04 | pass→pass | 9,134 | 8,817 | -3% | 1 | 1 | 0% | 1,619 | 1,161 | -28% | 0 | 0 | — |
case-05 | pass→fail | 6,374 | 5,196 | -18% | 1 | 1 | 0% | 1,268 | 651 | -49% | 0 | 0 | — |
case-06 | fail→fail | 9,958 | 2,504 | -75% | 1 | 1 | 0% | 1,665 | 658 | -60% | 0 | 0 | — |
case-07 | fail→pass | 9,108 | 6,691 | -27% | 1 | 1 | 0% | 1,482 | 1,494 | +1% | 0 | 0 | — |
case-08 | fail→fail | 6,400 | 2,695 | -58% | 1 | 1 | 0% | 1,089 | 589 | -46% | 0 | 0 | — |
case-09 | fail→pass | 6,038 | 3,216 | -47% | 1 | 1 | 0% | 1,030 | 788 | -23% | 0 | 0 | — |
case-10 | fail→pass | 9,140 | 3,091 | -66% | 1 | 1 | 0% | 1,599 | 746 | -53% | 0 | 0 | — |
case-11 | fail→pass | 6,641 | 2,734 | -59% | 1 | 1 | 0% | 1,120 | 698 | -38% | 0 | 0 | — |
case-12 | fail→pass | 6,096 | 1,950 | -68% | 1 | 1 | 0% | 969 | 563 | -42% | 0 | 0 | — |
case-13 | fail→pass | 10,800 | 3,796 | -65% | 1 | 1 | 0% | 1,810 | 913 | -50% | 0 | 0 | — |
case-14 | pass→pass | 5,179 | 3,148 | -39% | 1 | 1 | 0% | 856 | 753 | -12% | 0 | 0 | — |
case-15 | fail→fail | 7,639 | 2,426 | -68% | 1 | 1 | 0% | 1,278 | 642 | -50% | 0 | 0 | — |
case-20 | fail→fail | 2,069 | 5,693 | +175% | 1 | 1 | 0% | 301 | 549 | +82% | 0 | 0 | — |
case-16 | pass→pass | 2,675 | 2,327 | -13% | 1 | 1 | 0% | 464 | 670 | +44% | 0 | 0 | — |
case-17 | pass→pass | 4,398 | 6,186 | +41% | 1 | 1 | 0% | 839 | 1,204 | +44% | 0 | 0 | — |
case-18 | pass→pass | 11,386 | 3,613 | -68% | 1 | 1 | 0% | 1,903 | 783 | -59% | 0 | 0 | — |
case-19 | fail→fail | 10,814 | 6,700 | -38% | 1 | 1 | 0% | 1,897 | 570 | -70% | 0 | 0 | — |
case-21 | fail→fail | 4,542 | 5,142 | +13% | 1 | 1 | 0% | 715 | 624 | -13% | 0 | 0 | — |
case-22 | fail→pass | 6,145 | 2,372 | -61% | 1 | 1 | 0% | 1,061 | 518 | -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 16 counted toward the lift figure. The other 6 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 16 comparable cases. 2 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.