Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Triggers the WORK-PIPELINE when a user request starts with a [] tag (e.g., [new-feature], [bugfix], [WORK start]). Use this skill whenever you detect a [] tag at the beginning of a user message.
.claude/skills/davepoon-work-pipeline/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-09 | ✗→✓ | ▲ Improved | -39% | 0% |
| case-22 | ✗→✓ | ▲ Improved | 364% | 0% |
| case-19 | ✓→✗ | ▼ Worse | 10% | 0% |
| case-18 | ✓→✓ | ＝ Same ✓ | 12% | 0% |
| case-20 | ✓→✓ | ＝ Same ✓ | -8% | 0% |
When the user's message starts with a [] tag, start the WORK-PIPELINE by reading ../skills/sdd-pipeline/references/agent-flow.md and following the orchestration flow.
Any message starting with [...] triggers this pipeline:
[new-feature], [enhancement], [bugfix], [new-work], [WORK start]When this skill is triggered, Claude Code provides the "Base directory for this skill" as an absolute path. Derive the REFERENCES_DIR from it:
REFERENCES_DIR = {Base directory}/../sdd-pipeline/referencesYou MUST pass this absolute path to every sub-agent invocation (specifier, planner, scheduler, builder, verifier, committer). Include it at the top of the prompt text:
REFERENCES_DIR={absolute_path}Sub-agents need this path to read their reference files. Without it, they cannot find the files and will loop.
works/WORK-NN/Requirement.md, determines execution-mode (direct/pipeline/full)direct: call builder → committerpipeline: call builder → verifier → committer in sequencefull: call planner → ⛔ STOP for 2nd approval → scheduler → builder → verifier → committer] × NIf the user's message ends with "auto" or "자동으로", skip ALL approval steps and execute the entire pipeline automatically. This is the ONLY case where approval gates can be skipped.
User requirement: $ARGUMENTS
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 12,212 | 4,333 | -65% | 1 | 1 | 0% | 2,120 | 718 | -66% | 0 | 0 | — |
case-06 | fail→fail | 11,151 | 5,796 | -48% | 1 | 1 | 0% | 2,416 | 870 | -64% | 0 | 0 | — |
case-02 | fail→fail | 9,693 | 3,874 | -60% | 1 | 1 | 0% | 1,907 | 688 | -64% | 0 | 0 | — |
case-03 | fail→fail | 16,103 | 5,402 | -66% | 1 | 1 | 0% | 3,375 | 718 | -79% | 0 | 0 | — |
case-04 | fail→fail | 14,025 | 3,931 | -72% | 1 | 1 | 0% | 2,864 | 697 | -76% | 0 | 0 | — |
case-05 | fail→fail | 3,507 | 5,970 | +70% | 1 | 1 | 0% | 560 | 801 | +43% | 0 | 0 | — |
case-07 | fail→fail | 25,254 | 6,496 | -74% | 1 | 1 | 0% | 4,757 | 920 | -81% | 0 | 0 | — |
case-08 | fail→fail | 17,602 | 4,501 | -74% | 1 | 1 | 0% | 3,689 | 747 | -80% | 0 | 0 | — |
case-09 | fail→pass | 10,230 | 6,772 | -34% | 1 | 1 | 0% | 1,640 | 996 | -39% | 0 | 0 | — |
case-10 | fail→fail | 8,633 | 5,321 | -38% | 1 | 1 | 0% | 1,643 | 798 | -51% | 0 | 0 | — |
case-11 | fail→fail | 4,329 | 3,583 | -17% | 1 | 1 | 0% | 701 | 659 | -6% | 0 | 0 | — |
case-12 | fail→fail | 9,766 | 5,261 | -46% | 1 | 1 | 0% | 2,116 | 764 | -64% | 0 | 0 | — |
case-13 | fail→fail | 10,543 | 6,088 | -42% | 1 | 1 | 0% | 1,888 | 699 | -63% | 0 | 0 | — |
case-14 | fail→fail | 17,882 | 5,220 | -71% | 1 | 1 | 0% | 3,260 | 743 | -77% | 0 | 0 | — |
case-15 | fail→fail | 14,951 | 4,379 | -71% | 1 | 1 | 0% | 3,058 | 669 | -78% | 0 | 0 | — |
case-16 | fail→fail | 13,440 | 3,794 | -72% | 1 | 1 | 0% | 2,374 | 664 | -72% | 0 | 0 | — |
case-17 | fail→fail | 9,871 | 4,004 | -59% | 1 | 1 | 0% | 1,796 | 672 | -63% | 0 | 0 | — |
case-18 | pass→pass | 5,286 | 3,814 | -28% | 1 | 1 | 0% | 927 | 1,042 | +12% | 0 | 0 | — |
case-19 | pass→fail | 6,945 | 5,106 | -26% | 1 | 1 | 0% | 1,249 | 1,379 | +10% | 0 | 0 | — |
case-20 | pass→pass | 8,050 | 5,252 | -35% | 1 | 1 | 0% | 1,574 | 1,446 | -8% | 0 | 0 | — |
case-21 | fail→fail | 21,065 | 29,089 | +38% | 1 | 1 | 0% | 4,927 | 787 | -84% | 0 | 0 | — |
case-22 | fail→pass | 3,128 | 8,385 | +168% | 1 | 1 | 0% | 450 | 2,087 | +364% | 0 | 0 | — |
case-23 | fail→fail | 11,218 | 4,350 | -61% | 1 | 1 | 0% | 2,048 | 680 | -67% | 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. 23 cases were attempted, and 5 counted toward the lift figure. The other 18 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 +4 percentage points is the difference between those two pass rates over the 5 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.