Llm

As an experiment we wanted to add a feedback loop into LLM generated code via type annotations, docstrings, automated tests, and user feedback. We developed a simple package “pseudoscode” which uses the OpenAI API for generating code.