Synchronizes session data across multiple AI agents using a topological B3 Poset database structure.
Unified Reafference is a sophisticated cross-agent session management skill that enables seamless data persistence and state synchronization between Claude, Goose, and Amp. By leveraging mathematical frameworks like B3 Posets and Bumpus sheaves, it creates a 'topological chemputer' environment where agents can share context and feedback loops (reafference). This skill is essential for developers building complex, multi-agent orchestrations where maintaining a unified memory and state across different LLM instances is critical for consistency and logical flow.
Key Features
01Bumpus sheaves integration for topological data consistency
02Native support for DuckDB time-travel and narrative auditing
03Ergodic session handling for continuous agent coordination
042 GitHub stars
05Cross-agent session database synchronization for Goose, Claude, and Amp
06B3 Poset-based state management with Trit-based logic roles
Use Cases
01Maintaining a unified session history for auditing complex AI-to-AI interactions
02Managing state in multi-agent 'topological chemputer' research and experiments
03Synchronizing context between disparate AI agents during long-running development tasks
