Journal RAG

Enables semantic search through personal markdown journal entries using vector database technology for effortless retrieval of past experiences, thoughts, and activities.

The Journal RAG MCP server enables AI assistants to search through personal markdown journal entries using semantic similarity. It implements a vector database approach with ChromaDB and sentence-transformers to index journal content, supporting both full and incremental indexing of entries. The server exposes two main tools: one for querying journal entries based on semantic relevance, and another for updating the index when new entries are added. Built with GPU acceleration support for faster embedding generation on NVIDIA hardware, this implementation is particularly useful for users who maintain personal journals in markdown format and want to reference past experiences, thoughts, and activities during conversations with AI assistants without manually searching through entries.