@Bobbie Maloy Thanks @Aaron Kruger . This reminds me that I need to start making updates to some of my builds. No promises on all of them, but since this one has been mentioned a few times, I’ll go ahead and update it. I’ve since updated the HTML intake, as my newer PDF-to-MD process is a bit faster with larger files and now has the option to separate any embedded images. I also have to give credit to @Ari Evergreen for the original inspiration behind this approach. If I remember correctly, her implementation used a Python-based processing pipeline for the document conversion/extraction, which helped inspire the direction I took with mine. Her original post, “Stop feeding raw documents to your LLM,” is here: https://www.skool.com/cliefnotes/stop-feeding-raw-documents-to-your-llm?p=d9d73b5f I’ll see what I can do @Bobbie Maloy once I get a chance to get into my mobile office. The one within the-steward is good to go for simple files in the meantime, or just check out Ari’s.