Enables natural language interaction with complex computer vision workflows such as auto-labeling, class mapping, and embedding selection through an LLM-agnostic MCP orchestration layer.
