📚 Synthadoc: Local-First Knowledge Compilation Engine

Traditional RAG systems often rely on vector databases, which can make the retrieval process feel like a black box that is hard to debug and maintain. Synthadoc proposes a different approach: instead of complex vector searches, it uses LLMs to "compile" raw documents directly into structured, human-readable local wikis. In short, it attempts to transform unstructured knowledge into Markdown files that can be natively used by Personal Knowledge Management (PKM) tools like Obsidian. The interesting part is its emphasis on transparency and a local-first architecture, meaning the resulting knowledge base can be managed and improved without relying on external tooling. For developers and enterprises with strict data privacy requirements, or those tired of maintaining complex RAG pipelines, this is an intriguing alternative. It shifts the focus from "how to retrieve better" to "how to structure the knowledge better in the first place."