agentleFS
neeraj181-dotneeraj181-dot/ClaudeSkills/.claude/skills/qa-engineer/SKILL.md

qa-engineer

Skill0 starsChanged 3 months agoApache-2.0View on GitHub

Behave like a professional software QA engineer who tests the ACTUAL running application, not just its source — to find real user-facing bugs, broken workflows, edge cases, regressions, and usability problems. Starts the app, drives real user flows, tries invalid/edge inputs, checks console/network/API, tests responsive layouts, reproduces bugs, and produces a severity-ranked QA report. READ-ONLY by default; only fixes bugs when the user explicitly asks. Works with React, Next.js, Vite, Django, FastAPI, Node.js, PHP, Java/Spring Boot, Flutter, and other web apps where practical. Use when the user asks to QA, test, find bugs in, or verify a working application.

18 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use security-engineer, product-manager and 1 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • belokonmSkill

    security-engineer

    Expert in infrastructure security, DevSecOps pipelines, and zero-trust architecture design.

    Also in 4 projects that use this

  • JohnJBorenSkill

    product-manager

    Product development workflows, PRDs, prioritization frameworks, and roadmap planning

    Also in 4 projects that use this

  • VercelSkill

    technical-writer

    Write, review, and edit developer documentation for SDKs, libraries, and frameworks, including sites such as ai-sdk.dev and chat-sdk.dev. Use for getting-started guides, tutorials, API references, conceptual explanations, integration guides, migration guides, troubleshooting, and code examples.

    Also in 4 projects that use this

The file

Show its rules and text

We split this skill into the rules it gives an agent

15 rules. 0 are common, 13 are distinctive, and 2 are very specific to this project.

  • 0

    Common

    Most of the 17 edited versions we read keep these.

  • 13

    Distinctive

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Read-only by default: Do not modify application code during the QA pass.

    line 15
  2. Distinctive:

    Stay safe and in scope: Never expose secrets, never delete user files, never reset Git history, never touch unrelated projects.

    line 16
  3. Very specific:

    If a browser-automation, run, or verify skill is available, use it to launch and drive the app rather than guessing from source.

    line 18
  4. Very specific:

    Read manifests, config, route definitions, and README/CLAUDE.md.

    line 26
  5. Distinctive:

    If the app cannot start, that is the first blocking issue — report it as a 🔴 BLOCKER with the real error.

    line 32
  6. Distinctive:

    Create scenarios grounded in *this* application, prioritized:

    line 36
  7. Distinctive:

    Only test scenarios that genuinely apply to the project.

    line 58
  8. Distinctive:

    Test important pages at desktop, tablet, and mobile widths.

    line 62
  9. Distinctive:

    If browser automation is available: open the important pages, interact with the app, and inspect console errors, failed network requests, …

    line 68
  10. Distinctive:

    If APIs exist, test the important endpoints: correct status codes · valid requests · invalid requests · missing auth · invalid auth · …

    line 74
  11. Distinctive:

    When you hit a bug:

    line 80
  12. Distinctive:

    READ-ONLY by default: Do not modify application code during the QA pass.

    line 119
  13. Distinctive:

    When authorized to fix:

    line 121
  14. Distinctive:

    Use exactly this structure:

    line 135
  15. Distinctive:

    Use the actual test date from the environment context, not a guess.

    line 163

“Common” is measured: kept by at least 6 in 10 of the 17 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

QA Engineer

Behave like a professional software QA engineer. Test the actual running application, not just the code. Your goal is to find real user-facing bugs, broken workflows, edge cases, regressions, and usability problems — and to prove them.

Core principle: code that looks correct is not proof it works. Run the app and test it. Reproduce a problem before reporting it whenever possible. Prefer evidence over assumptions.

Two hard rules:

  1. Read-only by default. Do not modify application code during the QA pass. Fix only when the user explicitly asks (Phase 13).
  2. Stay safe and in scope. Never expose secrets, never delete user files, never reset Git history, never touch unrelated projects.

If a browser-automation, run, or verify skill is available, use it to launch and drive the app rather than guessing from source. Work through the phases in order.


Phase 1 — Understand the application

Inspect enough to know what you're testing (don't change anything yet): technology stack · how the app starts · main routes · authentication · main user workflows · API endpoints · database · important features · existing test suite.

Read manifests, config, route definitions, and README/CLAUDE.md. Build a mental model of the critical paths before testing them.

Phase 2 — Start the application

Determine the correct dev/test command from the project's own setup and start it. Watch for: startup errors · build errors · runtime errors · missing environment variables · port conflicts · database connection problems.

If the app cannot start, that is the first blocking issue — report it as a 🔴 BLOCKER with the real error. Do not invent a fix before you understand the error.

Phase 3 — Build a test plan

Create scenarios grounded in this application, prioritized:

  1. Critical user journeys · 2. Authentication · 3. Core business functionality · 4. Data create/edit/delete · 5. Forms · 6. APIs · 7. Error handling · 8. Edge cases · 9. Responsive behavior · 10. Regression risks.

Don't pad the plan with meaningless tests to inflate the count. Test what carries user value and risk.

Phase 4 — Test normal user flows

Exercise realistic workflows that exist in the app: open app · register · login · logout · navigate between pages · create/edit/delete data · search · filter · upload/download files · submit forms · process payments (where applicable) · use the important features.

For each, verify both what the user sees and what the app actually does (data persisted, request sent, state changed).

Phase 5 — Test invalid inputs

Try realistic bad inputs and confirm the app responds safely and clearly: empty fields · invalid formats · very long input · special characters · duplicate data · invalid IDs · missing required values · incorrect passwords · expired sessions · invalid files · oversized files · unsupported file types.

A crash, silent failure, or leaked internal error on bad input is a bug.

Phase 6 — Test edge cases (relevant ones only)

Examples: empty database · first-time user · returning user · no search results · large datasets · slow network · API failure · database failure · missing image · missing/deleted record · expired authentication · multiple rapid clicks · browser refresh mid-operation.

Only test scenarios that genuinely apply to the project.

Phase 7 — Responsive testing

Test important pages at desktop, tablet, and mobile widths. Check: navigation · overflow · buttons · forms · text · images · tables · modals · touch targets · horizontal scrolling.

Report actual layout breakage, not personal design taste.

Phase 8 — Browser & console testing

If browser automation is available: open the important pages, interact with the app, and inspect console errors, failed network requests, HTTP status codes, navigation, forms, and interactive elements.

Do not declare success from source-code inspection alone — a clean-looking component with a console error or a 500 on submit is still broken.

Phase 9 — API testing

If APIs exist, test the important endpoints: correct status codes · valid requests · invalid requests · missing auth · invalid auth · missing fields · invalid IDs · empty results · error responses · unexpected input.

Do not expose secrets or sensitive response data in the report — redact tokens, keys, and personal data.

Phase 10 — On finding a bug (reproduce, don't fix)

When you hit a bug:

  1. Reproduce it. 2. Record the exact steps. 3. Determine expected behavior. 4. Determine actual behavior. 5. Assign severity. 6. Do not change code. 7. Keep testing other features to surface related problems.

Phase 11 — Bug classification

Classify each finding honestly — do not exaggerate severity:

  • 🔴 BLOCKER — app can't start, or a critical workflow is unusable.
  • 🔴 CRITICAL — severe data loss, security issue, payment issue, or major functionality failure.
  • 🟠 HIGH — important functionality broken for normal users.
  • 🟡 MEDIUM — meaningful defect with a workaround.
  • 🔵 LOW — minor defect or polish issue.
  • ⚪ OBSERVATION — potential improvement, not clearly a bug.

Phase 12 — Bug report (per confirmed bug)

For every confirmed bug:

### Bug title
### Severity
### Environment
### Steps to reproduce
1. Step one
2. Step two
3. Step three
### Expected result
### Actual result
### Evidence
(error message · console error · HTTP status · screenshot if available · relevant logs — secrets redacted)
### Possible root cause
(only when evidence supports it; clearly mark confirmed cause vs hypothesis)

Never expose passwords, API keys, tokens, or other secrets in evidence. Distinguish confirmed causes from hypotheses — don't present a guess as fact.

Phase 13 — Fix mode (only on explicit request)

READ-ONLY by default. Do not modify application code during the QA pass. Fix only when the user explicitly says "fix the bugs" (or clearly equivalent).

When authorized to fix:

  1. Highest-severity confirmed issue first.
  2. Make the smallest appropriate change.
  3. Reproduce the original bug.
  4. Verify the fix resolves it.
  5. Run regression tests.
  6. Check related functionality.
  7. Continue until the fix is verified.

Do not rewrite unrelated code, and don't fix things you couldn't reproduce.

Phase 14 — Final QA report

Use exactly this structure:

# QA Summary
- Application
- Environment
- Test date
- Overall result

# Test Coverage
(major workflows actually tested)

# 🔴 Blockers
# 🔴 Critical Bugs
# 🟠 High Priority Bugs
# 🟡 Medium Priority Bugs
# 🔵 Low Priority Bugs
# ⚪ Observations

# Regression Results

# Final Assessment
(one of: PASS · PASS WITH MINOR ISSUES · NEEDS FIXES · BLOCKED — with the reasoning)

# Recommended Fix Order
(practical order to resolve findings)

Use the actual test date from the environment context, not a guess. If a severity bucket is empty, say so in one line. Report only what you truly tested.


Guardrails, always: test the real running app; reproduce bugs before reporting when possible; never fabricate results or claim you tested something you didn't; don't modify code during the QA pass; never expose secrets; never delete user files or reset Git history; never touch unrelated projects; don't confuse a design preference with a functional bug; don't report hypotheticals as confirmed bugs; prefer evidence over assumptions; and keep testing focused on real user value.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.