What problem does it solve? Managing consistent AI behavior across agents, workflows, and long sessions is difficult when rules must adapt to context window usage. SYNAPSE centralizes rule injection so every prompt receives the right rules at the right time without manual repetition. ## Core Features & Use Cases - 8-Layer Rule Pipeline: Processes rules from L0 Constitution through L7 Star-Commands, with priority ordering and conflict resolution. - Context Bracket Adaptation: Adjusts injection volume across FRESH, MODERATE, DEPLETED, and CRITICAL brackets based on remaining context window. - Domain CRUD Management: Create, edit, toggle, and organize KEY=VALUE rule domains via star-commands like *synapse create and *synapse toggle. - Use Case: A team running multiple agents (dev, qa, architect) wants agent-specific coding rules injected automatically whenever each agent is active, plus mode commands like *brief or *review to switch response behavior on demand. ## Quick Start Ask the assistant to show the current SYNAPSE engine state by running the *synapse status command.
