Automated prompt regression testing & assertion framework for LLMs in VS Code. Run .prompt.yaml tests, assert outputs, compare model responses, and track prompt drift.
Installation
Launch VS Code Quick Open (Ctrl+P), paste the following command, and press enter.
Unit testing, regression assertion, and eval framework for Large Language Model (LLM) prompts inside Visual Studio Code and Cursor IDE. Run .prompt.json and .prompt.yaml test files, assert output criteria, and catch prompt regressions before pushing to production.
Zero dependencies, instant local execution, works with online LLMs (Gemini, OpenAI, Claude, Ollama) and offline simulation.