Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →End-to-end feature work across src/, scripts/, tests/, manuscript, and docs for the Research Project Template. USE WHEN adding a pipeline-visible feature, new analysis stage, manuscript-facing output, or cross-layer integration — even without feature_addition prompt.
.claude/skills/docxology-template-feature-addition/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-10 | ✗→✓ | ▲ Improved | -10% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -33% | 0% |
| case-12 | ✗→✓ | ▲ Improved | -50% | 0% |
| case-17 | ✗→✓ | ▲ Improved | -63% | 0% |
| case-19 | ✗→✓ | ▲ Improved | -34% | 0% |
src/ or infrastructure/; script coordinates only.projects/<n>/scripts/ or root scripts/ stage hook if pipeline-visible.{{VARIABLE}} generated variables if numbers appear in prose../run.sh --project <n> --core-only slice.bashuv run python scripts/pipeline/stage_01_test.py --project <project> uv run python scripts/runner/execute_pipeline.py --project <project> --core-only
Checklist: references/checklist.md
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-04 | pass→pass | 9,521 | 8,482 | -11% | 1 | 1 | 0% | 1,959 | 2,013 | +3% | 0 | 0 | — |
case-10 | fail→pass | 9,928 | 4,482 | -55% | 1 | 1 | 0% | 1,252 | 1,132 | -10% | 0 | 0 | — |
case-01 | fail→fail | 3,663 | 4,511 | +23% | 1 | 1 | 0% | 170 | 648 | +281% | 0 | 0 | — |
case-02 | fail→fail | 3,761 | 4,191 | +11% | 1 | 1 | 0% | 194 | 570 | +194% | 0 | 0 | — |
case-03 | fail→fail | 17,981 | 5,229 | -71% | 1 | 1 | 0% | 313 | 777 | +148% | 0 | 0 | — |
case-05 | fail→fail | 3,761 | 5,964 | +59% | 1 | 1 | 0% | 561 | 1,411 | +152% | 0 | 0 | — |
case-06 | pass→pass | 19,973 | 3,614 | -82% | 1 | 1 | 0% | 3,919 | 815 | -79% | 0 | 0 | — |
case-07 | pass→pass | 10,585 | 3,533 | -67% | 1 | 1 | 0% | 1,855 | 1,034 | -44% | 0 | 0 | — |
case-08 | pass→pass | 9,796 | 2,917 | -70% | 1 | 1 | 0% | 1,781 | 924 | -48% | 0 | 0 | — |
case-09 | fail→pass | 10,077 | 3,634 | -64% | 1 | 1 | 0% | 1,489 | 1,004 | -33% | 0 | 0 | — |
case-11 | pass→pass | 8,775 | 2,284 | -74% | 1 | 1 | 0% | 1,518 | 724 | -52% | 0 | 0 | — |
case-12 | fail→pass | 8,741 | 2,121 | -76% | 1 | 1 | 0% | 1,620 | 804 | -50% | 0 | 0 | — |
case-13 | pass→pass | 7,963 | 2,067 | -74% | 1 | 1 | 0% | 1,256 | 705 | -44% | 0 | 0 | — |
case-14 | pass→pass | 10,934 | 4,756 | -57% | 1 | 1 | 0% | 1,699 | 1,225 | -28% | 0 | 0 | — |
case-15 | pass→pass | 5,269 | 2,225 | -58% | 1 | 1 | 0% | 791 | 746 | -6% | 0 | 0 | — |
case-16 | pass→pass | 9,737 | 7,888 | -19% | 1 | 1 | 0% | 1,738 | 1,830 | +5% | 0 | 0 | — |
case-17 | fail→pass | 11,533 | 1,754 | -85% | 1 | 1 | 0% | 1,853 | 690 | -63% | 0 | 0 | — |
case-18 | pass→pass | 10,744 | 3,453 | -68% | 1 | 1 | 0% | 1,922 | 955 | -50% | 0 | 0 | — |
case-19 | fail→pass | 8,294 | 3,221 | -61% | 1 | 1 | 0% | 1,504 | 992 | -34% | 0 | 0 | — |
case-20 | pass→pass | 10,404 | 3,255 | -69% | 1 | 1 | 0% | 1,788 | 973 | -46% | 0 | 0 | — |
case-21 | pass→pass | 4,373 | 2,266 | -48% | 1 | 1 | 0% | 704 | 695 | -1% | 0 | 0 | — |
case-22 | pass→pass | 14,637 | 4,467 | -69% | 1 | 1 | 0% | 2,221 | 1,102 | -50% | 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 19 counted toward the lift figure. The other 3 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 19 comparable cases.
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.