AI & LLM Usage

Read this carefully before spamming @Claude or whatever.

While I'm personally not a fan of using LLMs and AI, I think it greatly accelerates development of the project, especially in refactoring and atomic parts.

However, that doesn't mean you can just give an LLM full control and ask it to build features for you. Using LLM agents for automated contributions - in the sense of "@Gok, implement me X feature" - is strictly forbidden.

Not OK

• Asking AI to complete the whole process from start to finish.
• Not reviewing or testing the result and just opening a PR for it.
• Adding AI generative-image functionality to this project. This is a drawing app, not a Midjourney.
• Replying with an AI-generated reply.

OK

• Asking AI to "implement function A that takes inputs A, B, C and outputs D by these specific rules" - at the end of the day, writing code is not the useful job; making the final product is.
• Reviewing each line of the LLM output and fixing it yourself if needed.
• Taking full responsibility for the code an LLM made for you, and being OK to deal with it - even when asked to fix a bug an LLM can't fix.