Show HN: I built ThreadShelf to reuse hard-to-export AI chats like OpenRouter (github.com)

ThreadShelf has launched as a local-first tool designed to help users archive, search, and continue AI-generated conversations across multiple platforms like ChatGPT, Claude, and Google AI Studio. By implementing advanced local embeddings and a semantic search capability, ThreadShelf allows users to retrieve conversations not just by keywords but by meaning. This innovative platform normalizes output from different AI providers into a unified format and stores conversations in LanceDB, facilitating both convenience and control over data privacy. The significance of ThreadShelf lies in its ability to manage AI chat data more effectively, particularly as the volume of interspersed chat files across various platforms can be overwhelming. Its architecture allows conversation threads to be reconstructed and revisited, enhancing productivity for AI developers and researchers. With features such as a React UI, HTTP API access, and command-line interface, ThreadShelf supports a variety of workflows while maintaining user control through loopback operations for local processing. This initiative not only streamlines the workflow for users engaging with multiple AI models but also emphasizes data ownership and privacy in an era marked by increasing concerns over AI-generated content.