<critical_constraints> ❌ NO tests without edge cases ❌ NO generic assertions (assertTrue) → use specific (assertEqual, toEqual) ❌ NO test dependencies → each test must be independent ✅ MUST follow Arrange-Act-Assert (AAA) pattern ✅ MUST mock external dependencies (APIs, DB, filesystem) ✅ MUST use descriptive test names (what + expected outcome) </critical_constraints>

<detection> Detect framework from existing tests: - Python default: pytest - JS/TS default: Jest - Java: JUnit - Ruby: RSpec - Go: testing package </detection>

<test_categories>

•Normal cases: typical usage
•Edge cases: empty, zero, max, boundary
•Error cases: invalid input, exceptions
•Type edges: None/null, wrong types
•Performance: large inputs (if applicable) </test_categories>

<templates> ## Python (pytest) ```python class TestFunctionName: def test_normal_case(self): assert func(input) == expected

code

def test_edge_empty(self):
    assert func([]) == []

def test_error_raises(self):
    with pytest.raises(ValueError):
        func(invalid)

code


## JavaScript (Jest)
```javascript
describe('functionName', () => {
  it('should return X when Y', () => {
    expect(func(input)).toBe(expected);
  });
  
  it('should throw on invalid', () => {
    expect(() => func(null)).toThrow(TypeError);
  });
});

</templates> <heuristics> - Conditional logic → test both branches - Loop → test 0, 1, many iterations - Division → test divide by zero - Strings → test empty, unicode, long - Collections → test empty, single, many </heuristics>