Advise testing qa skills

Testing And QA Skill Advisor

Overview

Advisor for designing, evaluating, and improving AI agent skills in the testing & qa domain. Use this skill to turn a vague skill idea into a compact skill blueprint, not to perform the domain work itself.

When To Use

Use when a human asks to design, review, improve, migrate, publish, or validate an AI agent skill for testing & qa. Use when trigger quality, workflow shape, failure recovery, output format, references, scripts, or validation need improvement. Do not use when the human wants the agent to directly perform the testing & qa task instead of designing the skill.

Inputs And Tools

• User's skill idea, draft SKILL.md, target runtime, example prompts, and any existing references, scripts, or assets.
• Repository skill files, routing metadata, and validation reports when improving an existing repository.
• Read-only inspection first; ask before edits, publishing, running commands, or spawning agents.

Workflow

• Separate skill-design advice from target-domain execution.

1. Identify the target quality surface: unit, integration, end-to-end, browser, accessibility, performance, or release regression.
2. Define the test entry points and required environment setup.
3. Specify how the skill should select focused tests before broad suites.
4. Define evidence capture: logs, screenshots, traces, coverage, or failing assertions.
5. Define failure triage and retry rules.
6. Return a test-skill blueprint with validation examples.

Category Standards

• Keep the skill compact enough to trigger reliably and load without flooding context.
• Prefer reusable reference files and scripts for heavy domain-specific details.
• Tests must be tied to observable behavior.
• Skill must separate failing setup from failing product behavior.
• Browser skills must verify pages with screenshots or DOM evidence when possible.
• Performance skills must define thresholds.
• Accessibility checks must name the target standard or heuristic.
• Output must summarize command results clearly.

Verification

• Run at least two realistic prompts: one that should trigger the skill and one that should not.
• Check that the skill asks for missing inputs before giving brittle advice.
• Confirm failure modes include recovery behavior, not only warnings.
• Confirm validation commands or manual checks are safe for the target runtime.

Checklist

• Examples show concrete input and expected output.
• Verification can be repeated by a human or another agent.
• Trigger description names testing or QA tasks.
• Workflow identifies when to start or reuse dev servers.
• Skill defines how to handle flaky tests.
• Skill defines what evidence to collect.
• Failure modes cover missing dependencies, unavailable browsers, and timeouts.
• Output format includes test command, result, and next action.

Nice To Have

• Trigger collision notes for adjacent skills.
• PASS/PARTIAL/FAIL verification rubric.
• Reusable Playwright fixtures.
• Test data builders.
• Accessibility checklist.
• Performance budget reference.
• Screenshot artifact naming convention.

Example Skill Ideas

• Webapp testing skill.
• Regression triage skill.
• Accessibility audit skill.
• Performance smoke test skill.

Examples

Input: "Help me design a skill for testing & qa workflows that reviews drafts and flags missing evidence." Output: A compact skill blueprint with trigger description, ordered workflow, resources, failure recovery, validation prompts, and output format.

Input: "This testing & qa skill triggers too often and gives generic advice." Output: A revision plan that narrows triggers, moves heavy details to references, adds examples, and defines repeatable verification.

Failure Modes

• Test environment fails to start: capture logs and report setup blockers separately.
• Browser unavailable: suggest headless alternatives or manual verification.
• Flaky result: rerun once, compare evidence, and mark uncertainty.
• No tests exist: propose the smallest meaningful test before broader coverage.

Human Verification

Score the revised skill before calling it ready:

Discovery: PASS | PARTIAL | FAIL
Content: PASS | PARTIAL | FAIL
Verification: PASS | PARTIAL | FAIL
Critical gaps:
1. <gap>
Required fixes:
1. <fix>
Decision: READY | NEEDS REVISION

Output Format

Return:

Skill concept: <name>
QA surface: <surface>
Trigger description draft: <Use when...>
Workflow steps:
1. <step>
Evidence to collect:
- <evidence>
Failure recovery:
- <mode and recovery>
Validation prompts:
- <prompt>