Mnemora

Mnemora is a local-first, privacy-focused personal context engine that revolutionizes how you interact with your documents. Unlike cloud-based solutions, Mnemora processes everything locally using Ollama, ensuring your data never leaves your computer. The application supports multiple document types including Markdown (with Obsidian wiki-links, tags, and YAML frontmatter), PDFs (with full text extraction and page markers), plain text files, and code files (with language detection and structure extraction). It uses ChromaDB as a high-performance vector database for semantic search, allowing you to find information across thousands of documents using natural language queries. Every answer includes smart citations showing exactly which documents contributed to the response, along with relevance scores. The application features a modern, dark-themed interface built with React and Tailwind CSS, streaming responses for real-time AI generation, and is fully extensible as an open-source project.