Goldentrii/AgentRecall
🔌 MCP ServerGoldentrii
An MCP server providing persistent, context-aware memory for AI agents using the Intelligent Distance Protocol.
AgentRecall addresses the common limitation of stateless AI agents by providing a robust, persistent memory layer. Built as an MCP (Model Context Protocol) server, it enables agents to store and retrieve information across disparate sessions, ensuring that context is not lost when a conversation ends. The core innovation lies in its use of the Intelligent Distance Protocol, which optimizes memory retrieval by surfacing only the most contextually relevant data, rather than relying on simple keyword matching or exhaustive search. This ensures that agents remain efficient while maintaining a deep, evolving knowledge base of their interactions. The toolset includes granular control over memory lifecycles: 'session_start' and 'session_end' manage the temporal scope, while 'remember' and 'recall' handle the ingestion and retrieval of facts. A 'check' function allows for verification of stored state. By abstracting the complexities of vector storage and retrieval, AgentRecall allows developers to focus on building smarter, more personalized AI experiences that learn and adapt over time.
💡Highlights
- ├─Intelligent Distance Protocol
- ├─Persistent cross-session memory
- └─Five-tool MCP interface
🎯For
- ├─AI Application Developers
- └─Agent Framework Engineers