Local Vector Search for Jira, Confluence & Files (with MCP support)
A local tool that indexes documents from Jira, Confluence or local files into a vector database and searches them with hybrid vector and keyword search, keeping all data on the user's own machine.
This project indexes documents from Jira tickets, Confluence pages, or local files such as PDF, PPTX and DOCX into a local vector database, then searches them by hybrid retrieval combining vector search with BM25 keyword search merged through reciprocal rank fusion. All data stays on the user's machine except when exposed as an MCP tool to a remote AI agent.
Collections are updated incrementally, so only new or changed documents are re-indexed, and results can be filtered by metadata fields such as space, project or date. It uses SQLite with the sqlite-vec and BM25 extensions, ChromaDB, sentence-transformers, Unstructured and LangChain, and exposes search as an MCP tool for AI agents. It is written in Python and MIT licensed.