#ai-agents

Token-Efficient Memory: How Aura Caches BM25, Repo Maps, and Long-Term Context

Three layers of on device persistence that cut token waste and keep the AI focused between sessions.