Show HN: ThreadShelf – Obsidian and AI chats: vector search, MCP, llama.cpp

Category: ai-ml

Tags: local-first, vector-search, mcp, llama.cpp, ai-chat-archive, semantic-search

Score: 7.3/10 (Innovation: 7, Technical: 7, Documentation: 8, Utility: 7)

ThreadShelf is a local-first tool for archiving, semantically searching, and continuing AI conversations from multiple providers (ChatGPT, Claude, etc.) using local embeddings, LanceDB, and optional llama.cpp inference. It stands out by combining a unified normalized archive with semantic search, a web UI, CLI, HTTP API, and MCP server, while keeping data private. The ability to resume threads with local models or optionally external OpenRouter adds practical value.

Target audience: AI power users, developers, and privacy-conscious users who want to manage and search their AI conversation history locally

Repository: https://github.com/ChrystianSchutz/ThreadShelf · TypeScript · MIT · 15 stars

View on Hacker News