reversa-requirements

Generate requirements.md from Reversa state and extracted SDD evidence.

Updated May 13, 2026
One-click install
npx skills add https://github.com/cliffthelin/AXiomEngine --skill reversa-requirements-cliffthelin
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: reversa-requirements
Source: https://github.com/cliffthelin/AXiomEngine/tree/main/skills/reversa/reversa-requirements
Command: npx skills add https://github.com/cliffthelin/AXiomEngine --skill reversa-requirements-cliffthelin

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Reversa-requirements converts a free-form feature idea into a complete, evidence-anchored requirements document by using what was already extracted from the legacy codebase.

Core Features & Use Cases

  • Requirements generation from reverse-engineering artifacts: Writes requirements.md using the existing _reversa_sdd/ outputs (architecture, domain rules, inventory, and targeted code analysis).
  • Safe feature lifecycle management: Detects whether a previous feature is already in progress via physical artifacts and asks the user how to proceed instead of overwriting silently.
  • Hook execution integration: Runs or exposes .reversa/hooks.yml directives (before-requirements / after-requirements flow) to extend the pipeline without breaking the main workflow.
  • Quality-minded drafting: Supports doubt-marking with [DOUBT] (capped to avoid excessive uncertainty) and ensures the resulting document is ready for later plan/action steps.

Quick Start

Use the reversa-requirements skill to create requirements for a new feature by telling it what you want to build, for example: Create a new endpoint that summarizes customer purchases and returns totals by month, then generate the corresponding requirements document.

Frequently Asked Questions about reversa-requirements

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate requirements from legacy codebase reverse-engineering artifacts?

To generate requirements from legacy codebase artifacts, the Skill transforms free-form ideas into requirements.md anchored in existing _reversa_sdd/ outputs like architecture, domain rules, and inventory. It cites reverse-engineering evidence directly to produce auditable documentation.

How do I create an auditable requirements document for a new feature?

You create an auditable requirements document by providing a free-form feature idea, which the Skill converts into an atomic, evidence-cited requirements.md. It reads Reversa state and config to ensure the output is ready for downstream planning and quality steps.

Can I use feature planning documents without overwriting an in-progress feature?

Yes, you can safely manage feature planning without overwriting in-progress work because the Skill detects existing features via physical artifacts. It identifies if a previous feature is active and asks the user how to proceed rather than silently overwriting files.

How does intent resolution handle uncertainty when drafting requirements?

Intent resolution handles uncertainty by applying bounded [DOUBT] markers within the drafted requirements document. This caps excessive ambiguity and explicitly flags areas needing clarification, ensuring the output remains quality-minded and ready for later audit steps.

Does policy-driven development support custom hook execution during requirements generation?

Yes, policy-driven development supports custom hook execution by running or exposing .reversa/hooks.yml directives. It triggers before-requirements and after-requirements flows to extend the pipeline and integrate custom logic without breaking the main workflow.

What should I do if the requirements generation does not find relevant SDD evidence?

If requirements generation lacks relevant SDD evidence, the resulting requirements.md will contain bounded [DOUBT] markers indicating missing context. You should ensure the _reversa_sdd/ directory contains the necessary architecture, domain rules, and inventory outputs before drafting.