Frustrated by the complexity of existing RAG frameworks, the author built a lightweight, zero-dependency Node.js package that introduces an “agentic” feedback loop into retrieval workflows. Instead of a one-shot retrieval-and-answer pipeline, the system evaluates, rewrites, and retries queries until it finds relevant context or fails gracefully. The result is a more transparent, debuggable, and developer-friendly approach to building RAG systems—especially for rapid prototyping and smaller-scale applications.