Context

Generates missing tests for specified code and runs the test suite. Follows the project's testing standards and patterns.

Inputs

$ARGUMENTS - File paths to generate tests for, or "run" to just run existing tests

Steps

1. Determine Stack and Test Framework

Detect the stack from file extensions:

•.cs files → NUnit, Moq, FluentAssertions
•.ts/.tsx files → Vitest or Jest, React Testing Library
•.py files → pytest

Read the relevant standards file:

•standards/dotnet.md (Testing section)
•standards/typescript.md (Testing section)
•standards/python.md (Testing section)
•standards/testing.md (Cross-stack philosophy)

2. Analyze Code to Test

If $ARGUMENTS includes file paths (not just "run"):

•Read each file to understand its public API
•Identify the happy path and error paths
•Check for existing tests (look for corresponding test file)
•Identify dependencies that need mocking

3. Generate Tests

Follow these patterns per stack:

.NET (NUnit):

•Create test class in matching test project directory
•Use [TestFixture] with BaseTest if available
•Name: {ClassName}Tests.cs
•Methods: {Method}_When{Scenario}_Should{Expected} or {Method}_{Scenario}_{Expected}
•Use [SetUp] for mock creation
•Use Assert.EnterMultipleScope() for grouped assertions
•Test both success and error paths for Result<T> methods

TypeScript (Vitest/Jest):

•Create test file: {component}.test.tsx or {module}.test.ts
•Use describe blocks matching the component/function
•Use it with descriptive names
•For React components: use render, screen, fireEvent from Testing Library

Python (pytest):

•Create test file: test_{module}.py
•Use def test_{function}_{scenario}() naming
•Use pytest.fixture for setup
•Use pytest.raises for exception testing

4. Run Tests

Execute the test suite:

•.NET: dotnet test
•TypeScript: npm test or npx vitest run
•Python: pytest

5. Report Results

code

## Test Results

**Status:** PASS | FAIL
**Tests:** X passed, Y failed, Z skipped
**Coverage:** X% (if available)

### New Tests Created
- [file path] - [what it tests]

### Failed Tests (if any)
- [test name] - [failure reason]

Verification

•All new tests pass
•Tests cover both happy path and error paths
•Test names are descriptive and follow conventions
•No hardcoded dates or flaky patterns