spec-writer

To generate a product requirement document from a solution markdown file, you need a tool that reads solution.md, applies structural standards, and outputs a compliant spec.md. This process extracts version, scene, and functional narratives to ensure downstream development teams have clear, traceable requirements.

Structuring a spec.md file for downstream development requires applying scope definitions, layout and table conventions, and data dictionary references. This ensures the generated product requirements maintain structured table precision while keeping narratives natural for stakeholder approval.

Spec generation requires an existing solution plan before creating product requirements, specifically reading docs/plan/solution.md. You must finalize the solution content first, as the tool extracts version, scene, and functional narratives from this plan to build the compliant spec document.

You can use humanizer-zh post-processing to naturalize product requirement narratives during spec generation. This applies scope definitions and layout conventions to keep structured tables precise while ensuring the narrative text in your spec.md remains natural and readable.

Automated spec generation from solution plans should not be used when the solution.md content is incomplete or unconfirmed. The process relies on finalized solution plans to extract accurate version, scene, and functional narratives, so generating specs from drafts leads to non-compliant requirement documents.

Spec-writer handles data dictionaries and table layouts in spec.md files by applying validated layout and table conventions during generation. It reads the solution plan, enforces scope definitions, and references data dictionaries to ensure structured tables remain precise for downstream init-app workflows.