Aman

关于

Aman empowers developers by offering a powerful, privacy-first Local RAG (Retrieval Augmented Generation) solution specifically designed for querying codebases. It eliminates complex setup with its zero-configuration approach, allowing users to effortlessly search their private repositories. By combining advanced techniques like BM25, HNSW, and Tree-sitter, Aman provides highly accurate and contextually relevant results directly from your local environment, ensuring your data remains secure while enhancing your development workflow.

主要功能

• Zero-configuration setup for instant use
• Codebase understanding powered by Tree-sitter
• Supports integration with local LLMs (Ollama, MLX, Qwen3, Claude-code)
• Hybrid search capabilities (BM25, vector search, HNSW)
• Privacy-first local RAG for secure codebase queries
• 0 GitHub stars

使用案例

• Rapidly query and understand local code repositories without external dependencies
• Enhance developer productivity by instantly finding relevant code snippets and information
• Perform semantic searches across private project files while maintaining data privacy