Hey, I have build a RAG (Retrieval Augmented Generation) for RIOT, from the documentation and the example directory, that is I have divided the RIOT documentation/example directory into chunks and used an embedder to embed their semantic meaning into a vector, when a user has a request it retrieves the chunks that align most with this request, mostly you can use this to improve LLM’s responses when you have question about RIOT or if you want an LLM to give you RIOT code for a specific task, in my implemention at the end you have ready to go prompts which you can just paste into the LLM of your choice. I have tested this RAG with Mr. Hahm for a few examples and we saw some improvements, I would be very happy if you could try using this RAG and also report issues that you have with it. Here is the Repository with all necessary files: GitHub - KiyoLelou10/RAGforRIOTOS: This is a RAG Retrieval-Augmented Generation for the Operating System RIOT, it uses RIOT's documentation and also the examples directory retrieve relevant parts from code and documentation for a user query. These can be used to improve prompts to LLMs for generating RIOT code or explaining RIOT functionality