reversa-spec-sdd

Generate component-level SDD specifications from a Reversa PRD.

1.5k|383|Updated Apr 26, 2026
One-click install
npx skills add https://github.com/sandeco/reversa --skill reversa-spec-sdd-sandeco
Or copy as Structured Prompt for Agentβ–Ό
Please help me install this Agent Skill.
Skill: reversa-spec-sdd
Source: https://github.com/sandeco/reversa/tree/main/agents/reversa-spec-sdd
Command: npx skills add https://github.com/sandeco/reversa --skill reversa-spec-sdd-sandeco

SYSTEM DOCUMENTATION & REQUIREMENTS

πŸ’‘ This Skill includes references (resource) components.

What problem does it solve?

Converts PRDs into complete, traceable Spec-Driven Development (SDD) specs for AI-assisted implementation, reducing ambiguity and misalignment when moving from product requirements to executable agent-ready contracts.

Core Features & Use Cases

  • PRD-to-SDD decomposition: Breaks prd.md into logical components, one spec per component, based on the PRD’s scope, personas, and acceptance criteria.
  • Spec drafting with RFC-Pragmatic structure: Produces LLM-ready SDD documents that define what should happen (not how), including functional requirements and acceptance criteria.
  • Automated quality scoring and iteration: Scores each generated spec (0–100) against completeness, testability, clarity, scope, and edge-case coverage, then iterates up to defined limits.
  • Output ready for the forward pipeline: Writes specs under _reversa_sdd/sdd/ and finishes with a handoff to /reversa-forward, enabling the next step of the Reversa workflow.

Quick Start

Run /reversa-spec-sdd after /reversa-new so the skill reads _reversa_sdd/prd.md and generates _reversa_sdd/sdd/<componente>.md with quality scores.

Frequently Asked Questions about reversa-spec-sdd

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

FAQPage Schema
How do I convert a PRD into testable SDD specs for AI agents?β–Ό

To convert a PRD into testable SDD specs, you decompose product scope into logical components and generate one structured Spec-Driven Development markdown file per component with explicit acceptance criteria and quality scores.

What is spec-driven development decomposition for AI workflow?β–Ό

Spec-driven development decomposition breaks product requirements into logical parts, drafting LLM-ready functional specs that define what should happen rather than how, reducing ambiguity for AI-assisted implementation.

How are SDD specs scored for quality and agent readiness?β–Ό

SDD specs are scored 0–100 against completeness, testability, clarity, scope, and edge-case coverage, iterating automatically up to defined limits to ensure agent readiness before handoff.

Can I generate a single SDD spec from user-provided input without a full pipeline?β–Ό

Yes, you can generate a single SDD spec from user-provided input for standalone evaluation, or apply the skill within pipeline execution by reading a standard prd.md file to output multiple component specs.

What is the best way to structure acceptance criteria in an SDD markdown file?β–Ό

The best way to structure acceptance criteria in an SDD markdown file is using an RFC-Pragmatic structure with required RF IDs, explicit acceptance criteria, and 🟑 PLANJADO tagging for traceable agent-ready contracts.

What happens after generating SDD specs in the Reversa workflow?β–Ό

After generating SDD specs, the workflow finishes with a handoff to /reversa-forward, enabling the next step of the Reversa pipeline by writing specs under the designated sdd directory for downstream execution.