Concept Output Caching

What problem does it solve?

Cache concept outputs to eliminate redundant reads and accelerate rule evaluations across a single session, targeting sub-100ms access and high hit rates.

Core Features & Use Cases

• In-session memory cache (Level 1) provides fast reads of commonly accessed story, architecture, and implementation data.
• Automatic invalidation on file modification (mtime) to ensure data freshness.
• Optional persistent cache to survive across sessions (future enhancement) and batch cache statistics for monitoring.
• Seamless cache-aside integration with sync workflows to speed up repeated rule evaluations.

Quick Start

Enable the in-session cache and read concept outputs through the cache to accelerate subsequent reads during the current session.