reversa-pre-spec

Converts ideation decisions into a minimal pre-spec document for downstream specification pipelines.

1.5k|405|Updated Apr 26, 2026
One-click install
npx skills add https://github.com/sandeco/reversa --skill reversa-pre-spec
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: reversa-pre-spec
Source: https://github.com/sandeco/reversa/tree/main/agents/reversa-pre-spec
Command: npx skills add https://github.com/sandeco/reversa --skill reversa-pre-spec

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

After an ideation session ends with a decision, teams often jump straight into requirements or specs with too much detail or too little context. This Skill bridges that gap by converting the ideation decision into a minimal pre-spec package, with explicit open questions, so the next pipeline starts with exactly what it needs and nothing more.

Core Features & Use Cases

  • Minimal Scope Definition: Derives the smallest valuable first delivery slice from the session's decision and risk files, without writing requirements, data models, or architecture.
  • Explicit Doubt Tracking: Marks every unresolved point as [DÚVIDA] in a format consumable by the clarify step, instead of inventing answers.
  • Pipeline Routing: Resolves the correct next command based on project context, routing legacy projects to requirements gathering and greenfield projects to new-project ideation.
  • Use Case: After a brainstorm and arbitration session decides to build a notification module for a legacy system, run this Skill to produce a pre-spec.md capturing the problem, chosen path, minimum scope, non-goals, constraints, and open questions, then hand it off to the requirements pipeline.

Quick Start

Ask the agent to generate the pre-spec for the current ideation session after the decision file has been created by the arbiter step.

Frequently Asked Questions about reversa-pre-spec

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

FAQPage Schema
How do I turn an ideation decision into a specification input?

Run the pre-spec step after the arbiter produces decision.md. It reads the session files, asks at most three blocking questions, and writes a pre-spec.md containing the problem, chosen path, minimum scope, non-goals, constraints, and open doubts.

What is a pre-spec document in a specification workflow?

A pre-spec is a minimal handoff package between ideation and requirements. It states the problem, chosen path, smallest valuable delivery slice, explicit non-goals, active constraints, and unresolved questions, without containing requirements, data models, or architecture.

Does this Skill write requirements or technical design?

No. It explicitly refuses to write numbered requirements, data models, API contracts, architecture, estimates, or task plans. Those belong to downstream pipelines such as requirements, plan, or spec stages.

What happens if decision.md is missing from the session?

The agent stops and instructs you to run the arbiter step first. It requires framing.md, options.md, risks.md, and decision.md to exist in the session directory before generating the pre-spec.

How are unresolved questions handled in the pre-spec?

Every open point is marked as a [DÚVIDA] entry in the document body, in the same format the clarify step consumes. The agent never invents answers to close the document.

Can I use this for both legacy and greenfield projects?

Yes. It reads the context field from the active ideation state and routes legacy projects to the requirements pipeline and greenfield projects to the new-project pipeline, including legacy anchor sections only when relevant.