Which ready-made prompt templates does a QA engineer need?

Six basic ones: test cases (a table with positive/negative/edge/security types), a Jira bug report, a feature checklist, a regression checklist, a test plan, and test data. Copy, replace the placeholders in square brackets, add your project context at the top — and send.

How does Claude help prepare for a release?

With three checklists: a 30-minute smoke test (2-3 critical checks per module), post-deploy production checks (site opens, SSL is valid, a test payment goes through, logs are clean), and a Go/No-Go recommendation. Give it the test results — say, 112 passed out of 120 with 2 open majors — and you get a decision with the risks explained.

How do I extract testable requirements from a spec?

Paste the document and ask Claude to extract all testable requirements: for each one — ID, wording, check type, priority, and whether it can be automated. Then, from the requirements list, Claude builds a traceability matrix (requirement — test cases — coverage — status) so you can see gaps in coverage.

How do I test cross-browser compatibility and accessibility with Claude?

Ask for a checklist covering Chrome, Firefox, Safari, and Edge (last 2 versions) at resolutions from 1920x1080 down to 360x800: layout, fonts, forms, animations, modals. For accessibility — a separate WCAG 2.1 level AA checklist: contrast, keyboard navigation, screen reader support, visible focus, label-input association.

Can Claude help a QA engineer grow professionally?

Yes, in three roles: a mentor that explains techniques like boundary value analysis with concrete examples; an interviewer that asks 10 middle-QA-level questions and reviews your answers; and an automation teacher that writes your first Python + Selenium test and explains every line for someone with no programming background.

Claude for QA engineers: templates and advanced techniques

Contents

Ready-to-use prompt templates

Templates are prompts you adapt to your project. Copy, replace placeholders in square brackets, send to Claude. The foundation is the context formula from part one and the daily scenarios from part two.

Template: test cases

I'm testing: [feature description]
Platform: [web / mobile / API]
User: [user role]
Constraints: [limits, validations, business rules]

Write test cases in table format:
ID | Type | Title | Preconditions | Steps | Expected result | Priority

Types: positive / negative / edge / security
Priority: high / medium / low
Minimum 15 test cases.

Template: bug report

Format a bug report for Jira from this description:

[What happened — free text, however you want]

Format:
- Title (short, specific, starts with a verb)
- Severity: blocker / critical / major / minor
- Steps to reproduce (numbered)
- Expected result
- Actual result
- Environment: browser, OS, environment
- Frequency: always / sometimes / once
- Workaround (if any)

Template: feature checklist

Create a testing checklist for: [feature/page description]

Categories:
- [ ] Functionality (main scenario, alternatives)
- [ ] Validation (required fields, formats, limits)
- [ ] UI/UX (rendering, states, responsiveness)
- [ ] Security (access, injections, other users' data)
- [ ] Performance (large data, multiple clicks)

Template: regression checklist

Create a regression checklist for the release.

What changed: [list of tasks/PRs from the sprint]

System modules: [list main modules]

For each affected module:
- [ ] Smoke test (main scenario works)
- [ ] Related modules (what might break)
- [ ] Critical scenarios (payment, registration, login)

Set priority: what to check first.

Template: test plan

Build a test plan for the feature: [description]

Include:
1. Scope — what we test and what we DON'T test
2. Approach — types of testing (functional, UI, API, security)
3. Environments — where we test
4. Test cases — how many, which types
5. Entry/exit criteria
6. Risks — what could go wrong
7. Schedule — rough timeline

Template: test data

Generate test data for [form/API/table].

Fields: [list fields and types]
Constraints: [length, format, required]

Need:
- 5 fully valid sets
- 5 with invalid data (different errors)
- 5 with edge values
- 3 for security checks (XSS, SQL injection)

Format: [table / JSON / CSV / SQL INSERT]

Adapting to your project

Before using a template, add project context at the start:

Context: I'm testing a food delivery mobile app.
Platforms: iOS 16+, Android 12+.
Users: customer, courier, restaurant, admin.
Critical modules: order placement, payment, tracking.

[template below]

The more context — the more accurate the output.

Working with documentation

From spec — testable requirements

You have a 10-page document. Instead of reading and writing it down yourself:

Here's a technical spec (or paste the text).

Extract all testable requirements. For each:
- Requirement ID
- Wording
- Check type (functional / UI / performance / security)
- Priority (high / medium / low)
- Can it be automated (yes / no)

Traceability matrix

Here's the requirements list:
REQ-01: User can register via email
REQ-02: User can log in with email + password
REQ-03: Password reset via email
[...]

Create a traceability matrix:
Requirement | Test cases | Coverage | Status

Release prep

Smoke test

Create a smoke test checklist for [project name].

Main modules: [list]

For each module — 2-3 most critical checks.
The whole smoke test should take no more than 30 minutes.

Staging vs Production

Create a post-deploy production checklist:
- Site loads
- SSL certificate valid
- Login works
- Main business flow completes
- Payment system works (test transaction)
- Email notifications sent
- Logs free of critical errors

Go/No-Go

Here are pre-release test results:
- Total test cases: 120
- Passed: 112
- Failed: 5 (minor: 3, major: 2)
- Blocked: 3
- Open bugs: 2 major, 3 minor

Give a recommendation: Go or No-Go?
Explain why. What are the release risks?

Reporting

Summary for manager

Write a testing summary for the project manager.
Language: simple, no technical jargon.

Data:
- Sprint: Sprint 14
- Tasks: [list]
- Tests run: 85
- Bugs found: 12 (2 critical, 4 major, 6 minor)
- Blockers: payment doesn't work in Firefox
- Regression: no issues

Format: 5-7 sentences, key findings and recommendations.

Bug analysis

Here's a list of bugs from the last 3 sprints:
[paste list]

Find patterns:
- Which module is the buggiest?
- Which bug type is most frequent?
- Is there a trend — getting better or worse?
- Where should we strengthen testing?

Cross-browser, mobile, accessibility

Cross-browser testing

Create a cross-browser testing checklist for [page/feature].

Browsers: Chrome, Firefox, Safari, Edge (last 2 versions).
Resolutions: 1920x1080, 1366x768, 375x812 (iPhone), 360x800 (Android).

What to check in each browser:
- Layout not broken
- Fonts render
- Forms work
- Animations smooth
- Modals correct

Mobile specifics

What specific bugs occur in:
1. iOS Safari (latest)
2. Android Chrome
3. Mobile WebView (apps)

For each: a concrete bug example and how to check.

Accessibility (WCAG)

Create an accessibility checklist per WCAG 2.1 (level AA) for [page].

Categories:
- Text contrast
- Keyboard navigation
- Screen reader (alt text, aria-labels)
- Focus (visible, logical order)
- Forms (label-input pairing, error messages)
- Media (subtitles, controls)

Localization

Localization testing checklist:
- RTL languages (Arabic, Hebrew): layout not broken?
- Long strings (German): text doesn't get cut off?
- Date formats: DD.MM.YYYY vs MM/DD/YYYY
- Number formats: 1.000,50 vs 1,000.50
- Currencies: symbol before/after number
- Special chars in names: Ö, ñ, 日本語

Growth and development

Claude as a mentor

Explain in simple terms what boundary value analysis is.
Give an example for an "Age" field (18 to 65).
Which values to test and why?

What's the difference between smoke testing and regression testing?
When to use each? Give real project examples.

Interview prep

Ask me 10 questions as if it's a Middle QA interview.
After each answer — say what's good and what to add.
Topics: test design, API, SQL, processes, bug tracking.

First steps into automation

Even if you're not a programmer, Claude helps you understand:

Write a simple auto test in Python + Selenium for the scenario:
1. Open example.com
2. Enter login and password
3. Click "Sign in"
4. Verify the home page opened

Explain each line as if to someone who's never coded.

You don't have to automate right now. But understanding how it works makes you more valuable.

In the next part — your first contact with the project repository. Don't panic: Claude reads the code, not you.