nautilus-compass
View on GitHubReliability layer for multi-agent setups — keep agents coordinating without an orchestrator. Cross-dialog contracts + drift detection + black-box memory (no LLM extraction).
A local-first memory and reliability layer for AI agents, with hybrid retrieval, drift detection, and cross-agent coordination contracts. Provides MCP and A2A integrations for Claude Code and other agent clients.
Use Cases
Give AI agents persistent, searchable long-term memoryShare memory and coordination contracts across agentsDetect prompt drift against known failure patternsRetrieve relevant facts from past sessionsConnect agent clients through MCP or A2A
Built With
- Language
- Python
- Frameworks
- MCP · A2A · BM25 · Reciprocal Rank Fusion
Tags
agent memory · long-term memory · cross-agent · retrieval · drift detection · local-first · MCP · A2A · hybrid search · memory governance