Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Instantiate or update a workspace `UNITS.csv` from a selected pipeline and units template (deps/checkpoints/acceptance). **Trigger**: unit planner, generate UNITS.csv, execution contract, units template, 生成工作单元. **Use when**: 初始化/更新执行合同：从 pipeline 选择生成 `UNITS.csv`，或 scope 扩展需要新增 units/依赖。 **Skip if**: `UNITS.csv` 已正确反映当前 scope（无需重写）。 **Network**: none. **Guardrail**: 保持 CSV 合法；scope 增长时新增 units；只在满足验收后标 `DONE`。
.claude/skills/willoscar-unit-planner/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-02 | ✗→✓ | ▲ Improved | 45% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 12% | 0% |
| case-06 | ✗→✓ | ▲ Improved | -14% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -41% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -58% | 0% |
pipelines/*.pipeline.md) into a concrete UNITS.csv contract for a specific workspace.PIPELINE.lock.md (preferred) or a chosen pipelines/*.pipeline.mdtemplates/UNITS.*.csv referenced by the pipeline front matterUNITS.csv (in the workspace root)STATUS.md updated with next runnable unitsUses: templates/UNITS.*.csv.
PIPELINE.lock.md (or ask to create it via pipeline-router).units_template into the workspace as UNITS.csv if missing.UNITS.csv exists, only edit it to reflect:unit_id, skill, outputs, acceptance, checkpoint, status, owner.CHECKPOINTS.md; add missing checkpoints if the pipeline uses custom checkpoints.UNITS.csv parses as CSV and includes all required columns.depends_on references an existing U###.UNITS.csv, STATUS.md, CHECKPOINTS.md (for adding custom checkpoints).pipelines/ unless requested.PIPELINE.lock.md is missingFix:
pipeline-router (Mode A) or uv run python scripts/pipeline.py kickoff|init to create PIPELINE.lock.md before generating units.UNITS.csv becomes invalid CSV after editsFix:
inputs/outputs and avoid unescaped commas inside fields; validate with uv run python scripts/validate_repo.py.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-21 | fail→fail | 7,598 | 6,544 | -14% | 1 | 1 | 0% | 1,222 | 895 | -27% | 0 | 0 | — |
case-01 | fail→fail | 4,863 | 3,968 | -18% | 1 | 1 | 0% | 717 | 718 | +0% | 0 | 0 | — |
case-02 | fail→pass | 7,070 | 3,326 | -53% | 1 | 1 | 0% | 673 | 979 | +45% | 0 | 0 | — |
case-03 | fail→fail | 4,342 | 5,789 | +33% | 1 | 1 | 0% | 211 | 928 | +340% | 0 | 0 | — |
case-04 | fail→pass | 24,972 | 5,584 | -78% | 1 | 1 | 0% | 1,262 | 1,417 | +12% | 0 | 0 | — |
case-05 | pass→pass | 11,367 | 3,138 | -72% | 1 | 1 | 0% | 1,802 | 968 | -46% | 0 | 0 | — |
case-06 | fail→pass | 20,562 | 1,888 | -91% | 1 | 1 | 0% | 878 | 758 | -14% | 0 | 0 | — |
case-07 | fail→pass | 11,778 | 4,118 | -65% | 1 | 1 | 0% | 1,868 | 1,097 | -41% | 0 | 0 | — |
case-08 | fail→pass | 12,461 | 2,055 | -84% | 1 | 1 | 0% | 1,939 | 816 | -58% | 0 | 0 | — |
case-09 | fail→pass | 11,721 | 2,999 | -74% | 1 | 1 | 0% | 1,684 | 897 | -47% | 0 | 0 | — |
case-10 | fail→pass | 9,131 | 1,374 | -85% | 1 | 1 | 0% | 1,389 | 675 | -51% | 0 | 0 | — |
case-11 | fail→pass | 11,382 | 3,195 | -72% | 1 | 1 | 0% | 1,643 | 955 | -42% | 0 | 0 | — |
case-12 | fail→pass | 11,368 | 2,478 | -78% | 1 | 1 | 0% | 1,630 | 821 | -50% | 0 | 0 | — |
case-13 | pass→pass | 17,089 | 1,843 | -89% | 1 | 1 | 0% | 1,322 | 722 | -45% | 0 | 0 | — |
case-14 | fail→pass | 10,094 | 2,182 | -78% | 1 | 1 | 0% | 1,435 | 826 | -42% | 0 | 0 | — |
case-15 | pass→pass | 15,048 | 3,464 | -77% | 1 | 1 | 0% | 2,196 | 1,010 | -54% | 0 | 0 | — |
case-16 | fail→pass | 12,335 | 4,425 | -64% | 1 | 1 | 0% | 1,714 | 1,252 | -27% | 0 | 0 | — |
case-17 | pass→pass | 9,334 | 5,404 | -42% | 1 | 1 | 0% | 1,396 | 1,297 | -7% | 0 | 0 | — |
case-18 | pass→pass | 8,667 | 3,359 | -61% | 1 | 1 | 0% | 1,450 | 976 | -33% | 0 | 0 | — |
case-19 | fail→pass | 11,340 | 2,206 | -81% | 1 | 1 | 0% | 1,826 | 804 | -56% | 0 | 0 | — |
case-20 | fail→pass | 12,881 | 12,479 | -3% | 1 | 1 | 0% | 2,010 | 2,484 | +24% | 0 | 0 | — |
case-22 | fail→fail | 5,558 | 2,030 | -63% | 1 | 1 | 0% | 868 | 748 | -14% | 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 +59 percentage points is the difference between those two pass rates over the 17 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.