templates/skills/testing/aqa/SKILL.md
Design and maintain automated UI/API tests (framework patterns, stability, CI artifacts); not one-off manual checks
npx skillsauth add ozerohax/assistagents testing-aqaInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
<input_requirements> <required>Autotest goal and scenarios</required> <required>Environment and access</required> <required>Selected framework and versions</required> <required>Test data and cleanup strategy</required> <optional>Data and fixtures</optional> <optional>Stability requirements (flake rate)</optional> <optional>Artifacts requirements (screenshots/video/trace)</optional> <optional>Retry rules (if allowed)</optional> </input_requirements>
<design_rules> <rule importance="critical">Tests are independent and isolated</rule> <rule importance="critical">Each test has a clear success criterion</rule> <rule importance="high">Use stable selectors and patterns</rule> <rule importance="high">Assert business-meaningful states, not implementation details</rule> <rule importance="high">Minimize duplication via POM/fixtures</rule> <rule importance="high">Parameterize checks for variable data</rule> <rule importance="medium">Separate UI, API, and integration tests</rule> </design_rules>
<execution_rules> <rule importance="critical">Tests are deterministic and repeatable</rule> <rule importance="high">Failures are localized (clear failure reason)</rule> <rule importance="high">Flakes are fixed, not masked</rule> <rule importance="high">Collect artifacts on failure (screenshot/video/trace)</rule> <rule importance="medium">Tests should be fast and parallelizable</rule> <rule importance="medium">Retries are limited and documented</rule> </execution_rules>
<coverage> <focus> <item>Critical user paths</item> <item>Authorization and access control</item> <item>Core CRUD operations</item> <item>Errors and validation</item> </focus> </coverage><do_not> <item importance="critical">Do not use sleep instead of state-based waits</item> <item importance="high">Do not depend on unstable selectors</item> <item importance="high">Do not mix multiple scenarios in one test</item> <item importance="high">Do not make tests depend on execution order</item> <item importance="high">Do not share mutable state between tests</item> </do_not>
<example_patterns> <pattern>Page Object Model with isolated actions and assertions</pattern> <pattern>Fixtures for data and auth</pattern> <pattern>Explicit waits for UI state</pattern> <pattern>Selectors by role/data-testid/aria</pattern> </example_patterns>
testing
Visual regression (baseline diffs, UI stabilization)
testing
Design detailed reproducible test cases (Given/When/Then, one check per case); not smoke/regression checklists
development
Mobile testing (native/hybrid/web, gestures, offline, interruptions)
testing
Load testing (k6/JMeter scenarios, benchmarks)