PromptPex is an LLM-based tool to automatically generate and evaluate unit tests for a given AI model prompt. PromptPex extracts input and output specifications from the prompt and uses them to generate diverse, targeted, and valid unit tests. These tests are valuable in debugging and improving the prompt as well as understanding how the prompt performs when interpreted by different AI models.
