Ada
Self-hosted personal AI assistant with RAG memory, specialist plugins, and multi-modal capabilities for privacy-first conversation
Ada is a privacy-first personal AI assistant that runs entirely on your own hardware, providing long-term memory, web search, vision capabilities, and tool use without monthly API fees or cloud dependencies. Built by Luna System using Ollama for local LLM inference, ChromaDB for vector storage, and FastAPI for the backend, it features a specialist plugin system that enables contextual responses based on current media playback, OCR text extraction, DuckDuckGo web search, and other dynamic data sources. The ada-mcp component brings Ada's conversational AI and semantic memory search capabilities directly into code editors like VS Code, Cursor, Neovim, and Helix, enabling seamless AI assistance during development workflows while maintaining complete data privacy on local infrastructure.
Source
Repository: https://github.com/luna-system/ada/tree/HEAD/ada-mcp
Maintain Ada?
Let people know it's listed here — add the badge (live metrics, light/dark aware) or a plain link to your README or docs.
[Ada on getagentictools](https://getagentictools.com/mcp/luna-system-ada?ref=badge)