Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Run browser QA flows and write a JSON report
.claude/skills/elie222-qa-run/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-09 | ✗→✓ | ▲ Improved | -59% | 0% |
| case-16 | ✗→✓ | ▲ Improved | -11% | 0% |
| case-17 | ✗→✓ | ▲ Improved | -37% | 0% |
| case-20 | ✗→✓ | ▲ Improved | -67% | 0% |
| case-22 | ✗→✓ | ▲ Improved | -24% | 0% |
You are the browser QA flow orchestrator. Use the flow specs in qa/browser-flows to execute tests in a real browser.
Args: $ARGUMENTS
If no args or --help is present, print usage and stop.
Usage:
/qa-run --list/qa-run --all [--parallel] [--max-parallel=3]/qa-run --only=flow-a,flow-b [--parallel] [--max-parallel=3]/qa-run --group=api [--parallel]/qa-run --all --base-url=http://localhost:3000Base URL:
--base-url=<url> sets the Inbox Zero app URL (e.g. http://localhost:3000, https://www.getinboxzero.com, or any self-hosted URL).--base-url is NOT provided, ask the user which URL to test against before proceeding. Do not assume production or localhost.<base-url>/<account-id>/automation etc.Filtering:
--all or --only), only priority: high flows run. Low-priority flows are skipped.--all includes all flows regardless of priority.--only=flow-a,flow-b runs exactly the specified flows regardless of priority.--group=<name> filters to flows matching that group front matter value. Combinable with priority filtering.Process:
qa/browser-flows/README.md and the selected flow files.--list, print each flow id + title + group + priority + resources and stop.all, only, or default high-priority). Apply --group filter if present. Fail fast if any requested ids are missing.--parallel, batch flows so no batch contains overlapping resources, no flow lists another in conflicts_with (missing means none), and every flow in the batch has parallel_safe: true (missing means false).If batching is not possible, run sequentially.
qa/browser-flows/results/<run-id>.json and save screenshots underqa/browser-flows/results/<run-id>/.
qa/browser-flows/results/<run-id>.md following the template in the README.Output rules:
qa/browser-flows/README.md.failed and explain why.Behavior rules:
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-05 | pass→pass | 4,536 | 8,694 | +92% | 1 | 1 | 0% | 793 | 2,364 | +198% | 0 | 0 | — |
case-04 | pass→pass | 12,546 | 12,605 | +0% | 1 | 1 | 0% | 2,171 | 2,569 | +18% | 0 | 0 | — |
case-01 | fail→fail | 3,021 | 7,267 | +141% | 1 | 1 | 0% | 468 | 1,139 | +143% | 0 | 0 | — |
case-02 | fail→fail | 15,255 | 5,315 | -65% | 1 | 1 | 0% | 3,055 | 1,044 | -66% | 0 | 0 | — |
case-03 | fail→fail | 12,012 | 6,168 | -49% | 1 | 1 | 0% | 665 | 1,164 | +75% | 0 | 0 | — |
case-06 | pass→fail | 7,418 | 5,800 | -22% | 1 | 1 | 0% | 1,453 | 1,120 | -23% | 0 | 0 | — |
case-07 | pass→pass | 12,689 | 4,119 | -68% | 1 | 1 | 0% | 2,116 | 1,463 | -31% | 0 | 0 | — |
case-08 | pass→pass | 17,126 | 3,225 | -81% | 1 | 1 | 0% | 2,746 | 1,366 | -50% | 0 | 0 | — |
case-09 | fail→pass | 20,127 | 3,189 | -84% | 1 | 1 | 0% | 3,198 | 1,314 | -59% | 0 | 0 | — |
case-10 | pass→pass | 11,461 | 4,340 | -62% | 1 | 1 | 0% | 1,860 | 1,522 | -18% | 0 | 0 | — |
case-11 | pass→pass | 12,764 | 3,879 | -70% | 1 | 1 | 0% | 1,753 | 1,418 | -19% | 0 | 0 | — |
case-12 | pass→pass | 9,086 | 3,443 | -62% | 1 | 1 | 0% | 1,402 | 1,403 | +0% | 0 | 0 | — |
case-13 | pass→pass | 7,805 | 2,753 | -65% | 1 | 1 | 0% | 1,206 | 1,128 | -6% | 0 | 0 | — |
case-18 | fail→fail | 9,226 | 1,489 | -84% | 1 | 1 | 0% | 1,540 | 1,030 | -33% | 0 | 0 | — |
case-14 | pass→pass | 9,570 | 2,604 | -73% | 1 | 1 | 0% | 1,328 | 1,233 | -7% | 0 | 0 | — |
case-15 | pass→pass | 11,498 | 3,472 | -70% | 1 | 1 | 0% | 1,797 | 1,304 | -27% | 0 | 0 | — |
case-16 | fail→pass | 8,932 | 2,855 | -68% | 1 | 1 | 0% | 1,540 | 1,365 | -11% | 0 | 0 | — |
case-17 | fail→pass | 9,701 | 2,347 | -76% | 1 | 1 | 0% | 1,700 | 1,064 | -37% | 0 | 0 | — |
case-19 | fail→fail | 13,132 | 3,437 | -74% | 1 | 1 | 0% | 2,099 | 1,330 | -37% | 0 | 0 | — |
case-20 | fail→pass | 18,173 | 1,814 | -90% | 1 | 1 | 0% | 3,188 | 1,060 | -67% | 0 | 0 | — |
case-21 | pass→pass | 6,326 | 2,337 | -63% | 1 | 1 | 0% | 996 | 1,128 | +13% | 0 | 0 | — |
case-22 | fail→pass | 11,212 | 2,771 | -75% | 1 | 1 | 0% | 1,539 | 1,172 | -24% | 0 | 0 | — |
case-23 | pass→pass | 13,433 | 3,647 | -73% | 1 | 1 | 0% | 2,164 | 1,364 | -37% | 0 | 0 | — |
case-24 | pass→pass | 11,259 | 2,778 | -75% | 1 | 1 | 0% | 1,653 | 1,160 | -30% | 0 | 0 | — |
case-25 | pass→pass | 11,130 | 2,799 | -75% | 1 | 1 | 0% | 1,827 | 1,325 | -27% | 0 | 0 | — |
case-26 | pass→pass | 15,170 | 4,153 | -73% | 1 | 1 | 0% | 2,293 | 1,399 | -39% | 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. 26 cases were attempted, and 23 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 +15 percentage points is the difference between those two pass rates over the 23 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.