Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Generate Zephyr test cases from Jira stories: parse acceptance criteria and business rules, impact-analyze existing TCs (update vs. create new), and draft correctly named test cases. Use for AC-to-test generation; defer post-generation Jira linking and manual test-case quality review.
.claude/skills/hoangnguyen0403-quality-engineering-zephyr-test-generation/SKILL.md| Model | Eval pass | Runs |
|---|---|---|
| gemini-3.6-flash | 100% | 11 |
| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 34% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 24% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -17% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 63% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 0% | 0% |
?expand=renderedFields — HTML authoritative for platform colors:#00B8D9 = Web · #36B37E = Mobile · #FF991F = Web+Mobile
quality-engineering-test-plan-authoring scenario expansion: ultra-short AC, no expected behavior, bundled ACs, contradiction. Ask before drafting.Get Issue Link Test Cases with Jira issue key (e.g., {PROJECT}-{ID}).[Module] and [Screen] keywords + check sibling issue links.zephyr_test_plan.md before writing.Create Test Case (with customFields included — no separate Update needed) → Create Test Case Steps → Create Test Case Issue LinkGet Test Case Steps, show before/after diff, wait for explicit approval, then Update Test Case.| AC row | Action | | ------------------------------------------- | --------------------------------------------------------------- | | Single row [ WEB + MOBILE ] | ONE TC, Platform = "Web and Mobile", no platform prefix in name | | Two rows same behavior, different platforms | TWO TCs with Web_ / Mobile_ prefix — never merge |
Web_ / Mobile_ only when platform-exclusive; omit prefix for Web and Mobile.["Client user", "Client admin", "Internal sales rep", "External sales rep"]. not default to Client user.Web_ or Mobile_ prefix for platform-exclusive TCs — copy verbatim from artifact draft; omit prefix only when Platform = "Web and Mobile"."folder": {"id": X} in all PUT payloads.zephyr_test_plan.md before each run."System works" → "Banner 'Success' is visible".ASSUMED and surface it for review.When this skill applies, preserve the following domain terminology or equivalent concrete examples in the answer when relevant:
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-15 | pass→pass | 10,937 | 11,767 | +8% | 1 | 1 | 0% | 1,715 | 2,044 | +19% | 0 | 0 | — |
case-01 | fail→pass | 32,212 | 32,604 | +1% | 1 | 1 | 0% | 4,460 | 5,971 | +34% | 0 | 0 | — |
case-02 | fail→pass | 33,679 | 30,763 | -9% | 1 | 1 | 0% | 4,716 | 5,864 | +24% | 0 | 0 | — |
case-03 | fail→fail | 11,679 | 66,034 | +465% | 1 | 1 | 0% | 852 | 4,970 | +483% | 0 | 0 | — |
case-04 | pass→pass | 12,043 | 8,365 | -31% | 1 | 1 | 0% | 1,223 | 1,588 | +30% | 0 | 0 | — |
case-05 | fail→pass | 12,922 | 8,747 | -32% | 1 | 1 | 0% | 1,914 | 1,593 | -17% | 0 | 0 | — |
case-16 | pass→pass | 9,709 | 11,233 | +16% | 1 | 1 | 0% | 1,476 | 1,988 | +35% | 0 | 0 | — |
case-06 | fail→pass | 13,823 | 11,750 | -15% | 1 | 1 | 0% | 1,242 | 2,029 | +63% | 0 | 0 | — |
case-07 | fail→fail | 11,719 | 6,276 | -46% | 1 | 1 | 0% | 1,661 | 2,001 | +20% | 0 | 0 | — |
case-08 | pass→pass | 15,762 | 11,319 | -28% | 1 | 1 | 0% | 1,573 | 1,982 | +26% | 0 | 0 | — |
case-09 | fail→pass | 11,018 | 3,649 | -67% | 1 | 1 | 0% | 1,685 | 1,683 | -0% | 0 | 0 | — |
case-22 | pass→pass | 16,894 | 21,953 | +30% | 1 | 1 | 0% | 1,787 | 3,590 | +101% | 0 | 0 | — |
case-10 | pass→pass | 11,692 | 8,292 | -29% | 1 | 1 | 0% | 831 | 1,465 | +76% | 0 | 0 | — |
case-11 | fail→fail | 13,435 | 9,381 | -30% | 1 | 1 | 0% | 1,195 | 1,702 | +42% | 0 | 0 | — |
case-12 | fail→fail | 13,300 | 5,349 | -60% | 1 | 1 | 0% | 1,244 | 1,833 | +47% | 0 | 0 | — |
case-13 | fail→pass | 6,259 | 11,586 | +85% | 1 | 1 | 0% | 807 | 2,232 | +177% | 0 | 0 | — |
case-14 | pass→pass | 8,976 | 8,196 | -9% | 1 | 1 | 0% | 1,044 | 1,536 | +47% | 0 | 0 | — |
case-17 | pass→pass | 18,112 | 4,185 | -77% | 1 | 1 | 0% | 1,662 | 1,721 | +4% | 0 | 0 | — |
case-18 | fail→pass | 19,655 | 3,241 | -84% | 1 | 1 | 0% | 2,125 | 1,497 | -30% | 0 | 0 | — |
case-19 | fail→fail | 9,102 | 3,593 | -61% | 1 | 1 | 0% | 494 | 1,589 | +222% | 0 | 0 | — |
case-20 | pass→pass | 18,298 | 18,672 | +2% | 1 | 1 | 0% | 2,085 | 3,210 | +54% | 0 | 0 | — |
case-21 | pass→pass | 15,681 | 20,089 | +28% | 1 | 1 | 0% | 1,785 | 3,822 | +114% | 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 +32 percentage points is the difference between those two pass rates over the 22 comparable cases. 1 case got worse with the skill loaded, and it is 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.