Show HN: MindBase – an LLM that maintains a wiki from your notes and papers (github.com)

MindBase has introduced an innovative AI research assistant that creates and manages a markdown-based wiki from user-contributed sources like papers and articles. Unlike traditional AI solutions that require active re-derivation of answers, MindBase utilizes an LLM to automate wiki maintenance by reading input, identifying contradictions, and producing structured entries. This allows users to access synthesized information directly from the wiki, enhancing knowledge retention and organization. Significantly, MindBase distinguishes itself from conventional note-taking tools and retrieval-augmented generation (RAG) systems by compounding knowledge rather than merely storing it. As users contribute new insights or documents, the AI enriches the existing content, ensuring that the context evolves over time. The architecture comprises multiple layers including source storage, user-generated content, and a dynamic context file that is continually updated. With early access available for beta testing, MindBase promises to revolutionize how researchers interact with their notes, transforming passive data into an evolving repository of synthesized insights.