菜单
Run Playwright/Cucumber tests by provided name or grep pattern. Use when the user asks to "run tests", "run a test", "execute tests", "run a scenario", or wants to trigger the test suite.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
Automate browser interactions, test web pages and work with Playwright tests.
Create a new Cucumber BDD scenario in features/qavajs.feature for the SauceLabs demo app. Use when the user asks to "create a test", "add a scenario", "write a test case", or describes a new flow to test.
Diagnose and fix a failing Playwright/Cucumber test scenario. Use when the user asks to "fix a test", "fix a failing test", "repair a scenario", or a test is broken and needs investigation.
| name | run-test |
| description | Run Playwright/Cucumber tests by provided name or grep pattern. Use when the user asks to "run tests", "run a test", "execute tests", "run a scenario", or wants to trigger the test suite. |
| argument-hint | scenario name or grep pattern |
| allowed-tools | ["Bash"] |
Run Playwright/Cucumber e2e tests for the SauceLabs demo app.
$ARGUMENTS
Determine what to run based on the arguments:
npm run test
npx playwright test --project=agent --grep "<pattern>"
where <pattern> is the scenario name or keyword provided by the user.
Report the results concisely:
report/ and traces to test-results/For each failed scenario, also print a bug report block in this exact format (ready to paste):
**Scenario:** <scenario name>
**Steps:**
<numbered list of all Gherkin steps in the scenario, exactly as written>
**Expected result:**
<what the last step or assertion was expecting — derive from the failing step text>
**Actual result:**
<the actual error or failure message from the test output>
Derive the steps list from the feature file source (read features/*.feature) so the wording matches the spec, not the log output.