概要
Paparats empowers AI coding assistants by providing deep, semantic understanding of your entire codebase, whether it's a single repository or a multi-project workspace. It indexes code at function/class boundaries, embeds it using local Jina Code Embeddings via Ollama, and stores it in a private Qdrant vector database. This enables AI assistants to perform context-aware searches, answering complex questions like 'where is the rate limiting logic?' with exact, relevant code snippets, all while ensuring your code never leaves your machine. With real-time file watching, language-aware chunking, and advanced query expansion, Paparats optimizes both search quality and token efficiency.
