deep-research-prompt-architect

To design prompts for agentic LLM workflows, use an interactive, research-driven process that surfaces ambiguities and gathers requirements before structuring a programmatic delegation payload for final generation.

The best way to structure requirements before drafting LLM prompts is through interactive analysis that clarifies needs, surfaces edge cases, and eliminates ambiguity. This captures thorough frontmatter and workflow decisions for reproducible outputs.

Yes, you can use this interactive workflow to improve existing prompt specifications. It applies to refining existing prompts by orchestrating research, structured questioning, and frontmatter preparation before final drafting.

Yes, this prompt design workflow supports codebase examination and domain APIs. Its research-driven synthesis incorporates web-domain knowledge, codebase examination, and domain APIs as needed to structure the delegation payload.

To choose between implementation paths for an LLM skill, the workflow presents a two-workflow option design with a recommended option identified upfront, ensuring you select an optimal path before final prompt generation.

You should not use an interactive research workflow for prompt generation when your requirements are fully defined with zero ambiguity. This process specifically targets eliminating ambiguity and surfacing edge cases through structured questioning before drafting.