gsd-spec-phase

Run Socratic interviews to score ambiguity and write SPEC.md.

4|1|Updated Mar 22, 2026
One-click install
npx skills add https://github.com/lgwanai/openforce --skill gsd-spec-phase-lgwanai
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: gsd-spec-phase
Source: https://github.com/lgwanai/openforce/tree/main/skills/gsd-spec-phase
Command: npx skills add https://github.com/lgwanai/openforce --skill gsd-spec-phase-lgwanai

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps you clarify what a workflow phase must deliver, by turning vague inputs into falsifiable requirements using structured Socratic questioning and quantitative ambiguity scoring.

Core Features & Use Cases

  • Socratic interview loop: Runs up to 6 rounds with rotating perspectives to uncover missing context before committing decisions.
  • Quantitative ambiguity scoring: Scores ambiguity across 4 weighted dimensions after each round to measure progress toward clarity.
  • SPEC.md generation as a workflow gate: Writes a validated SPEC.md only when ambiguity is low enough, ensuring discuss-phase can proceed with stable “what/why” requirements.
  • Codebase scouting first: Reads existing project context (e.g., PROJECT.md, REQUIREMENTS.md, ROADMAP.md, STATE.md) and surveys the repository state before asking questions.

Quick Start

Run the skill for your phase by calling openforce /gsd-spec-phase "spec-phase --auto" to produce a SPEC.md that discuss-phase can use on the next run.

Frequently Asked Questions about gsd-spec-phase

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

FAQPage Schema
How do I turn vague project requirements into falsifiable specifications?

To generate a SPEC.md for your phase, run the skill via the command interface with your phase specified, allowing it to scout the codebase and conduct the Socratic interview automatically. It writes the document once the ambiguity score reaches the required threshold.

What is the ambiguity scoring threshold for passing a specification gate?

The ambiguity scoring threshold for passing a specification gate requires a score of 0.20 or less across four weighted dimensions. This gating rule ensures that downstream phases receive stable, validated requirements before solution planning begins.

How many rounds of Socratic questioning are needed to clarify project scope?

Up to 6 rounds of Socratic questioning with rotating perspectives are needed to clarify project scope and uncover missing context. After each round, quantitative ambiguity scoring measures progress toward the required clarity threshold.

Does generating a SPEC.md require existing repository context like ROADMAP.md?

Generating a SPEC.md does not strictly require existing repository context, but the skill reads available files like PROJECT.md, REQUIREMENTS.md, ROADMAP.md, and STATE.md first. Surveying this repository state helps tailor the Socratic interview to your project.

What happens if the ambiguity score never drops below 0.20?

If the ambiguity score never drops below 0.20, the specification gate remains closed and the SPEC.md file is not written. This prevents downstream solution planning from proceeding with unstable or incomplete requirements.