STAR: Semantic Temporal Associative Retrieval - A Local-First Graph-Based Context Engine

STAR (Semantic Temporal Associative Retrieval) is a local-first, graph-based information retrieval system designed to enable resource-constrained devices to navigate large-scale personal knowledge corpora. Unlike traditional dense vector retrieval systems that require loading complete indices into RAM, STAR implements a sparse bipartite graph approach that retrieves only relevant 'atoms' of information required for a given query.