Provides compression, retrieval, and statistics for local context-economy when interacting with GPT/Codex, enabling efficient token usage and exact recovery of compacted content.
