AI/LLMs are permitted to be used with this project. With that said, there are some guidelines that must be followed when using LLMs. Issues and pull requests that do not follow these guidelines may be closed without review or explanation.
- Using an LLM to generate content must be disclosed. This does not include using an LLM solely for analysis, such as understanding of the codebase.
- You are responsible for all contributions. This includes ensuring that the content does not violate intellectual property rights.
- All communication must be written by a human. If we wanted to use LLMs in an automated way, including communication, we can do it ourselves; we neither need nor want you to do this. Small snippets for context are acceptable but must be clearly indicated as originating from an LLM.
- All contributions must be reviewed in their entirety by a human. If you cannot explain any given part of your contribution, including reasoning, it is not acceptable.
- LLMs must not be included as an author or co-author in commits. Ensure that your editor does not do this automatically.
That is not to say that LLMs are not useful. They are great for rote tasks such as generating boilerplate code and tests, but they are not a substitute for human reasoning and understanding.
For reporting of security vulnerabilities, please see SECURITY.md. You must provide a human-written description of the vulnerability and how to reproduce it; the reproduction may be generated by an LLM.