metadata
language:
- en
- id
license: mit
tags:
- humanoid
- instruction-clarity
- pre-execution
- human-interaction
- reasoning
- llm
instruction-clarity-controller
Model Description
instruction-clarity-controller is a language model designed to assess the clarity and executability of natural language instructions before action.
Rather than executing tasks, the model determines whether an instruction is clear enough to proceed, requires clarification, contains contradictions, is incomplete, or poses safety risks. It is intended to improve communication between humans and humanoid or agent systems.
Intended Use
- Human–humanoid instruction clarification
- Pre-execution instruction assessment
- AI agent communication management
- Safety-aware interaction layers
Output Format
The model outputs JSON only:
{
"label": "VALID | AMBIGUOUS | CONTRADICTORY | INCOMPLETE | UNSAFE",
"confidence": 0.0
}