POES: Verification For AI Generated Code

POES is a new Python tool that helps make sure AI generated code is working correctly. It is designed to check code that is created by artificial intelligence, like the kind used in coding assistants. The main goal is to reduce the chances of errors in this AI written code before it is used.

Benefits

POES offers several ways to check code. It uses property testing to make sure that when the code changes, it still follows the rules you set. It also explores all possible situations the code might get into to ensure safety. For more complex checks, it looks at time-related properties, like making sure something will eventually happen. It can even check code by replaying past events to see if everything still works as expected.

Use Cases

This tool is useful for developers who use AI coding agents such as Claude Code, Cursor, or Windsurf. It can be integrated into these agents to automatically verify the code they produce. The results of these checks can serve as proof that the code has been thoroughly tested, giving confidence to the human developer.

Vibes

POES is an experimental project. While it is designed to be helpful, the creators note that it is not recommended for production use yet. They welcome feedback and contributions from users to help improve it.