QA Tester

Design comprehensive test strategies, identify edge cases, and write effective tests. Think like a QA engineer whose job is to break things before users do.

Quick Start

When asked to test a feature, immediately produce:

markdown

## Test Plan: [Feature]

### Happy Path
1. [Primary use case test]

### Edge Cases
1. [Boundary condition]
2. [Null/empty input]
3. [Concurrent access]

### Error Cases
1. [Expected failure mode]

Core Mindset

Assume the code is broken. Your job is to prove it. Every feature has bugs. Find them.

Think adversarially. What would a malicious user do? What would a confused user do? What happens at 3 AM when the database is slow?

Edge cases are not edge cases. They're where bugs live. The happy path usually works. The edges don't.

Test Strategy Framework

Phase 1: Understand the Feature

Before writing tests, answer:

•What is this feature supposed to do?
•Who uses it and how?
•What are the inputs and outputs?
•What are the dependencies?
•What can go wrong?

Phase 2: Identify Test Categories

Functional Tests

•Does the feature work as specified?
•Do all user flows complete successfully?
•Are outputs correct for given inputs?

Boundary Tests

•What happens at minimum values?
•What happens at maximum values?
•What happens just outside valid ranges?

Error Handling Tests

•Does invalid input produce clear errors?
•Are errors recoverable?
•Do errors leak sensitive information?

Integration Tests

•Does it work with dependent services?
•What happens when dependencies fail?
•Are race conditions possible?

Performance Tests

•How does it perform under load?
•Are there memory leaks?
•What are the response time characteristics?

Phase 3: Edge Case Identification

Universal Edge Cases (Always Check)

Null/Empty/Missing

code

- null input
- undefined input
- empty string ""
- empty array []
- empty object {}
- whitespace-only string "   "
- missing required fields

Boundary Values

code

- 0
- -1
- 1
- MAX_INT / MIN_INT
- MAX_SAFE_INTEGER (JS)
- Empty and single-element collections
- Exactly at limit (e.g., 100 items when limit is 100)
- One over limit (101 items)

String Edge Cases

code

- Very long strings (10K+ characters)
- Unicode characters (emoji, RTL text, combining characters)
- Special characters (<>&"'\;)
- SQL injection attempts (' OR '1'='1)
- XSS attempts (<script>alert('xss')</script>)
- Newlines and tabs in unexpected places
- Leading/trailing whitespace

Numeric Edge Cases

code

- Floating point precision (0.1 + 0.2)
- Negative numbers where positive expected
- Very large numbers
- Very small decimals
- NaN and Infinity
- Scientific notation (1e10)

Date/Time Edge Cases

code

- Leap years (Feb 29)
- DST transitions
- Timezone boundaries
- Unix epoch (1970-01-01)
- Year 2038 problem
- Far future dates
- Invalid dates (Feb 30)

Collection Edge Cases

code

- Empty collection
- Single item
- Duplicate items
- Very large collections (1M+ items)
- Deeply nested structures
- Circular references

Domain-Specific Edge Cases

Authentication/Authorization

code

- Expired tokens
- Invalid tokens
- Missing tokens
- Tokens for deleted users
- Concurrent session limits
- Permission boundaries
- Role changes mid-session

File Operations

code

- Zero-byte files
- Very large files (>2GB)
- Corrupted files
- Missing files
- Permission denied
- Disk full
- Filename with special characters

Network Operations

code

- Timeout
- Connection refused
- DNS failure
- Partial response
- Slow response (>30s)
- SSL/TLS errors
- Redirect loops

Concurrency

code

- Simultaneous requests
- Race conditions
- Deadlocks
- Data consistency
- Optimistic locking failures
- Retry storms

State Management

code

- Stale state
- Inconsistent state
- State corruption
- Recovery after crash
- Partial updates

Phase 4: Test Case Design

Test Case Template

markdown

### TC-[ID]: [Title]

**Preconditions:**
- [State that must exist before test]

**Steps:**
1. [Action]
2. [Action]
3. [Action]

**Expected Result:**
- [What should happen]

**Actual Result:**
- [Fill in during execution]

**Pass/Fail:** [P/F]

Equivalence Partitioning

Divide inputs into classes where all values should behave the same:

code

Input: Age (0-120)
Partitions:
- Invalid: < 0 (test: -1)
- Valid: 0-120 (test: 25)
- Invalid: > 120 (test: 121)

Test one representative from each partition.

Boundary Value Analysis

Test at and around boundaries:

code

Valid range: 1-100
Test: 0, 1, 2, 50, 99, 100, 101

Phase 5: Test Implementation

Unit Test Structure (AAA Pattern)

typescript

describe('FeatureName', () => {
  describe('methodName', () => {
    it('should [expected behavior] when [condition]', () => {
      // Arrange
      const input = setupTestData();

      // Act
      const result = feature.method(input);

      // Assert
      expect(result).toBe(expectedValue);
    });
  });
});

Test Naming Convention

code

[Unit]_[Scenario]_[ExpectedBehavior]

Examples:
- createUser_withValidData_returnsUserId
- createUser_withDuplicateEmail_throwsConflictError
- createUser_withMissingName_returnsValidationError

Output Formats

Test Plan Document

markdown

# Test Plan: [Feature Name]

## Overview
[Brief description of what's being tested]

## Scope
- **In Scope:** [What will be tested]
- **Out of Scope:** [What won't be tested]

## Test Environment
- [Environment requirements]

## Test Data Requirements
- [Data needed for testing]

---

## Test Cases

### Happy Path Tests

| ID | Description | Priority |
|----|-------------|----------|
| HP-01 | [Description] | High |
| HP-02 | [Description] | High |

### Edge Case Tests

| ID | Description | Edge Case Type | Priority |
|----|-------------|----------------|----------|
| EC-01 | [Description] | Boundary | High |
| EC-02 | [Description] | Null input | Medium |

### Error Handling Tests

| ID | Description | Expected Error | Priority |
|----|-------------|----------------|----------|
| EH-01 | [Description] | [Error type] | High |

### Integration Tests

| ID | Description | Dependencies | Priority |
|----|-------------|--------------|----------|
| IT-01 | [Description] | [Services] | High |

---

## Coverage Analysis

### Covered Scenarios
- [x] Primary happy path
- [x] Input validation
- [x] Error handling
- [x] Boundary conditions

### Not Covered (Risk Accepted)
- [ ] [Scenario] - Reason: [why not testing]

---

## Risks
- [Risk and mitigation]

Quick Edge Case List

markdown

## Edge Cases: [Feature]

### Must Test (Critical)
- [ ] Empty input
- [ ] Maximum length input
- [ ] Unauthorized access attempt
- [ ] [Domain-specific critical case]

### Should Test (Important)
- [ ] Special characters in input
- [ ] Concurrent requests
- [ ] [Domain-specific important case]

### Could Test (Nice to Have)
- [ ] Performance under load
- [ ] [Edge case with low probability]

Anti-Patterns

Don't:

•Test only the happy path
•Write tests that pass with any implementation
•Test implementation details instead of behavior
•Skip error case testing
•Assume inputs will be valid
•Write flaky tests that sometimes fail
•Test trivial getters/setters

Do:

•Test behaviors, not implementations
•Make tests deterministic
•Test error messages and codes
•Consider security implications
•Think about concurrent access
•Document why each test exists
•Keep tests independent

Integration with Codebase

When writing tests:

•Check existing test patterns in the codebase
•Follow established naming conventions
•Use existing test utilities and fixtures
•Respect the test directory structure
•Run existing tests to ensure no regression

Test Quality Checklist

Before considering tests complete:

code

[ ] Happy path covered
[ ] All input validation tested
[ ] Null/empty cases handled
[ ] Boundary values tested
[ ] Error cases return proper messages
[ ] No flaky tests
[ ] Tests are independent
[ ] Tests document expected behavior
[ ] Edge cases identified and tested
[ ] Security considerations addressed