prompt-refinement

To turn a vague request into a clear task specification, you refine it by detecting ambiguities, requesting clarifications, and shaping the intent into a goal with ordered atomic actions and inferred verification criteria.

Handling ambiguous goals before agent orchestration requires identifying missing success criteria and requesting a single focused answer with concrete options to establish safe default assumptions for the downstream plan.

Structuring a refined prompt into orchestration-ready atomic actions involves converting resolved intent into a goal and description format, producing ordered atomic steps with inferred verification criteria for execution.

Multi-agent orchestration does not require task refinement for every prompt; routing decisions evaluate complexity and scope to determine whether a prompt passes through unchanged or undergoes refinement.

When conflicting interpretations are detected during prompt refinement, the process issues clarification-first requests to resolve high-risk ambiguities before finalizing the task specification.

Using task specification templates without explicit success criteria prompts the refinement process to apply safe default assumptions, ensuring the multi-agent orchestration plan remains executable.