Lately I've been experimenting with automated AI Code Reviews using Github Actions triggered as part of our pull request process.

The idea is not to replace human verifications & reviews but rather have developers focusing on more important aspects of pull requests such as security concerns, overall functionality and knowledge sharing instead of trivial things like spell-checking, naming conventions and unused variables.

Found this really nice action on the GitHub Marketplace called ”AI Code Review Action”, which after just a few minutes of configuration was up and running and gave some really nice feedback on initial testing.

Spell-cheking:

Empty comments:

Unused variables:

So far this is just in experimental phase but initial findings seems good.

Cheers! ❤️