Local-first semantic knowledge graph with magnetic-pull retrieval

Local-first semantic knowledge graph enables users to build and query a personal, offline knowledge base using natural language, with a novel "magnetic-pull" retrieval system that surfaces related concepts based on semantic proximity. It is designed for researchers, writers, and knowledge workers who want private, fast, and context-aware data management without relying on cloud services. The project is interesting because it combines the autonomy of local-first software with the intelligence of semantic graphs, offering a new paradigm for personal knowledge organization that feels intuitive and responsive.