Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Why-What-Acceptance backlog format that connects every work item to strategic business objectives, with INVEST quality gates and observable acceptance criteria.
.claude/skills/borghei-wwas/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 27% | 0% |
| case-13 | ✗→✓ | ▲ Improved | 182% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 153% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 106% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 11% | 0% |
Create backlog items using the Why-What-Acceptance (WWAS) format. This format ensures every piece of work connects to strategic context, includes a concise description that serves as a "reminder of the discussion" rather than a detailed specification, and defines high-level acceptance criteria focused on observable outcomes.
The format has three parts:
Before an item enters a sprint, it must pass the INVEST gates (Independent, Negotiable, Valuable, Estimable, Small, Testable).
job-stories/ instead.Before writing the item, confirm these inputs. If any is unknown or vague, ASK — do not assume:
Stop rule: ask only the 2-3 that most change the output. If the user says "just draft it," proceed and list your assumptions at the top of the artifact.
Load the reference that matches the task; keep this file lean and pull detail on demand.
assets/wwas_template.md — ready-to-use WWAS templates.job-stories/ for situation-driven stories focused on user context rather than strategic alignment.brainstorm-okrs/ to define the objectives that WWAS items connect to.summarize-meeting/ to capture refinement discussions that inform the What.../jira-expert/ for ticket creation with structured fields.In Scope: Writing items in WWAS format, applying INVEST gates, connecting work to strategic objectives, facilitating refinement, converting existing items to WWAS, integrating with Jira.
Out of Scope: Situation-driven requirements (job-stories/), ideation/discovery (discovery/brainstorm-ideas/), OKR definition (execution/brainstorm-okrs/), detailed technical specs, sprint planning/capacity (../scrum-master/).
Limitations: WWAS adds most value with clearly defined objectives (OKRs, North Star). Without strategic context the Why becomes a forced exercise. It fits product/feature work better than pure tech-debt/infra items. Teams transitioning from user stories may need 2-3 sprints to build fluency.
| Integration | Direction | What Flows | |-------------|-----------|------------| | job-stories/ | Complementary | Job stories add situational context (When); WWAS adds strategic context (Why). Use both for complete requirements | | summarize-meeting/ | Meetings -> WWAS | Refinement discussions produce the What; decisions produce acceptance criteria | | ../jira-expert/ | WWAS -> Jira | WWAS items become Jira tickets with structured description fields | | execution/brainstorm-okrs/ | OKRs -> WWAS | Team OKRs provide the strategic objectives that Why statements reference | | execution/prioritization-frameworks/ | WWAS -> Prioritization | WWAS items scored via RICE or other frameworks for backlog ordering | | discovery/brainstorm-ideas/ | Ideas -> WWAS | Validated ideas decompose into WWAS backlog items with strategic traceability |
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 13,467 | 9,332 | -31% | 1 | 1 | 0% | 2,188 | 2,776 | +27% | 0 | 0 | — |
case-13 | fail→pass | 7,758 | 13,803 | +78% | 1 | 1 | 0% | 1,155 | 3,258 | +182% | 0 | 0 | — |
case-02 | fail→fail | 11,845 | 8,712 | -26% | 1 | 1 | 0% | 2,008 | 2,507 | +25% | 0 | 0 | — |
case-03 | fail→fail | 15,849 | 9,684 | -39% | 1 | 1 | 0% | 2,506 | 2,679 | +7% | 0 | 0 | — |
case-04 | pass→fail | 4,898 | 9,273 | +89% | 1 | 1 | 0% | 797 | 2,741 | +244% | 0 | 0 | — |
case-05 | pass→pass | 12,404 | 10,986 | -11% | 1 | 1 | 0% | 1,960 | 2,895 | +48% | 0 | 0 | — |
case-06 | pass→fail | 31,023 | 12,294 | -60% | 1 | 1 | 0% | 5,773 | 3,057 | -47% | 0 | 0 | — |
case-07 | fail→pass | 6,097 | 7,987 | +31% | 1 | 1 | 0% | 969 | 2,448 | +153% | 0 | 0 | — |
case-08 | fail→pass | 9,593 | 12,154 | +27% | 1 | 1 | 0% | 1,467 | 3,025 | +106% | 0 | 0 | — |
case-09 | pass→pass | 13,744 | 10,492 | -24% | 1 | 1 | 0% | 2,086 | 2,657 | +27% | 0 | 0 | — |
case-10 | fail→pass | 10,264 | 4,873 | -53% | 1 | 1 | 0% | 1,712 | 1,892 | +11% | 0 | 0 | — |
case-11 | pass→pass | 12,586 | 11,219 | -11% | 1 | 1 | 0% | 1,962 | 2,852 | +45% | 0 | 0 | — |
case-12 | fail→pass | 5,366 | 8,537 | +59% | 1 | 1 | 0% | 818 | 2,442 | +199% | 0 | 0 | — |
case-14 | fail→pass | 9,444 | 10,705 | +13% | 1 | 1 | 0% | 1,376 | 2,728 | +98% | 0 | 0 | — |
case-15 | pass→pass | 10,783 | 8,845 | -18% | 1 | 1 | 0% | 1,595 | 2,624 | +65% | 0 | 0 | — |
case-16 | fail→pass | 9,143 | 9,485 | +4% | 1 | 1 | 0% | 1,462 | 2,608 | +78% | 0 | 0 | — |
case-17 | fail→fail | 12,502 | 13,128 | +5% | 1 | 1 | 0% | 1,933 | 3,133 | +62% | 0 | 0 | — |
case-18 | fail→pass | 5,365 | 10,720 | +100% | 1 | 1 | 0% | 762 | 2,715 | +256% | 0 | 0 | — |
case-19 | fail→pass | 12,140 | 11,281 | -7% | 1 | 1 | 0% | 2,021 | 2,800 | +39% | 0 | 0 | — |
case-20 | pass→pass | 9,801 | 9,058 | -8% | 1 | 1 | 0% | 1,594 | 2,625 | +65% | 0 | 0 | — |
case-21 | fail→fail | 13,082 | 14,305 | +9% | 1 | 1 | 0% | 1,984 | 3,421 | +72% | 0 | 0 | — |
case-22 | fail→fail | 14,864 | 5,896 | -60% | 1 | 1 | 0% | 2,190 | 2,047 | -7% | 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. The headline lift of +36 percentage points is the difference between those two pass rates over the 22 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.