generate-tests

Criteria-driven mode generates tests by parsing provided spec files or task IDs containing acceptance criteria. It produces a complete set of failing tests for your new feature, ensuring implementation aligns directly with the specified requirements.

Yes, you can automatically create characterization tests from existing source code using the code-analysis mode. By providing file paths, the system analyzes the code behavior and generates corresponding tests to capture the current logic before refactoring.

Yes, the test generation tool fully supports Python, TypeScript, and JavaScript projects. It automatically detects the project's test framework, generating appropriate pytest, Jest, or Vitest test files based on your specific environment.

Test files are generated efficiently for large codebases by spawning parallel test-writer agents. This concurrent processing approach analyzes multiple files or acceptance criteria simultaneously, significantly reducing the total time required to produce tests.

The best way to start practicing TDD with automated test generation is to provide your acceptance criteria in a spec file. The system generates failing tests based on those specs, allowing you to write implementation code until the tests pass.