📚 Nocturne Memory: Graph-Structured Memory for MCP Agents

This is not just another vector database. Nocturne Memory is a lightweight, rollbackable, and visual long-term memory server designed specifically for MCP agents. It aims to replace traditional vector RAG with persistent, graph-like structured memory. The direction here is clear: traditional vector retrieval often struggles with context retention and complex relationships over time. Nocturne attempts to solve this by maintaining a structured memory graph that persists across different models, sessions, and tools. It also positions itself as a drop-in replacement for OpenClaw. If you are building agents that require long-term context and are frustrated with the limitations of standard RAG, this graph-based memory approach is worth exploring.