Context intelligence layer for LLM agents: persistent memory with write-time dedup, sensitivity tagging, conflict detection, and hierarchical decay. ~12ms. No LLM calls. MIT.
↑ +1 today
Discover the best Prompt Caching open source repositories on GitHub. Explore trending tools, live star counts, source code, and developer projects updated daily.