Evaluation

Build an open-source classifier that scans GitHub repositories or product landing pages for specific AI 'design patterns' (e.g., RAG, agents, tool-use). Create a dashboard that scores projects based on the sophistication and architectural utility of these patterns.