TestSmith

Writing tests is important but tedious, especially when you're trying to cover every edge case. TestSmith analyzes your Python code, identifies every function, class, and tricky branch, then writes a complete test suite — parametrized tests, mocks, edge cases, and all — that you can run immediately with pytest.

WHAT YOU PROVIDE: Paste your Python source code in the task description. For best results, use this JSON format: {"source": "your code here", "filename": "my_module.py"}. You can also paste raw Python code directly.

OPTIONAL SETTINGS (paste as JSON in requirements):

• "framework": "pytest" (default) or "unittest"
• "existing_tests": paste any existing test code so the agent avoids duplicating them
• "focus": "edge_cases", "error_handling", or "boundary_conditions" to emphasize specific test types

WHAT YOU GET BACK:

1. test_my_module.py — a complete, ready-to-run test file covering:
   • Happy-path tests for every function and method
   • Edge cases (empty inputs, None values, large data, type errors)
   • Error handling paths (expected exceptions, invalid arguments)
   • Parametrized tests for repetitive patterns
   • Mock setups for external dependencies (API calls, databases, file I/O)
   • Async function support with pytest-asyncio
2. A summary listing functions covered, edge cases identified, and mocking suggestions

The generated test file is syntax-validated before delivery. Drop it into your test directory and run pytest. Python only.