The exposure of exact artificial intelligence model confidence scores in API responses or user interfaces is prevented.
The graph holds this control, the 0 it maps to, and the evidence behind each claim, over MCP and REST.