forge

To create an XML prompt for Claude Opus or Codex CLI, use an interview-driven prompt forge that guides you through structured questions and outputs a tagged prompt with role, mission, context, requirements, workflow, tooling, validation, and done sections.

An interview-driven prompt is a structured blueprint generated through adaptive questioning to reduce ambiguity in complex prompt design. You need it for long-horizon, tool-heavy coding tasks or agent system prompts where unstructured requests cause back-and-forth.

To write a system prompt for coding agents with XML tags, run an interview-driven prompt forge that produces a complete XML-style prompt with defined sections, plus an optional SPEC.md handoff for requirements gathering before implementation.

Yes, this prompt engineering workflow supports GPT-5.5 and Codex CLI by generating standardized XML-tagged prompts that work across multiple harnesses, ensuring your coding agent receives a consistent, ready-to-run prompt blueprint.

The best way to gather requirements before a long coding task is using an automated, structured interview that surfaces requirements and outputs an XML prompt with an optional SPEC block, reducing ambiguity and back-and-forth in complex prompt design.

Your coding agent prompt lacks structure for tool-heavy workflows because it is missing standardized XML-tagged sections for tooling, validation, and done criteria, which an interview-driven prompt forge generates by guiding you through adaptive questions.