clarification-enforcer

To stop an agent from making assumptions on ambiguous user input, use a clarification enforcement mechanism that halts execution and asks specific, binary preference questions to resolve uncertainty before proceeding.

Handling missing context in complex agent workflows requires enforcing proactive questioning. The agent identifies contextual gaps, flags missing details crucial for task completion, and prompts the user for explicit requirements.

An agent proceeds with conflicting information instead of asking for clarification when it lacks a guardrail to detect ambiguity. Adding a clarification enforcer forces the agent to halt and formulate targeted questions to resolve the conflict.

To prompt users for specific requirements when they give vague requests like improve the code, implement ambiguity detection. This flags the vague request and forces the agent to ask whether improvement means performance, readability, or new features.

Yes, you can enforce explicit user-guided decision-making for every task execution by applying a clarification guardrail across all agent interactions. This ensures the agent halts whenever precise input is missing.

You should not use a clarification guardrail when task execution requires autonomous action without user input or when workflows are designed to handle errors dynamically without halting to ask binary preference questions.