How does the 'ask, don't assume' rule work?

The skill is explicitly forbidden from inventing requirements. If a design document is vague or missing details, the skill will stop and ask the user for clarification before proceeding.

What kind of output does this skill produce?

It generates a centralized specs/spec.xml file and a details directory containing modular XML files for specific domains like data models, API endpoints, and authentication.

How does this fit into the Arcforge workflow?

Refining is a prerequisite for the planning stage. Once the spec is finalized and committed, it provides the necessary context for arc-planning to generate implementation tasks.

What is the main purpose of the arc-refining skill?

It acts as a bridge between high-level design and implementation by converting informal documents into structured, authoritative XML specifications that guide downstream AI tasks.

Can I skip the clarification questions?

Yes, while the agent is designed to be thorough, the user can say 'enough, continue' at any time to bypass the iterative refinement phase.

Arc Spec Refiner

Name: Arc Spec Refiner
Author: GregoryHo

byGregoryHo

0•

Productividad y Flujo de Trabajo

Transforms high-level design documents into structured, authoritative XML specifications with rigorous validation and acceptance criteria.

The Arc Spec Refiner is a critical component of the Arcforge workflow engine designed to eliminate ambiguity in the AI development lifecycle. It converts informal design drafts into high-fidelity, structured XML specifications that serve as the single source of truth for all downstream coding tasks. By enforcing a strict 'ask, don't assume' policy and utilizing an iterative clarification process, this skill ensures that every requirement has clear acceptance criteria, identified dependencies, and traceable mappings, preventing the AI from inventing requirements during implementation.

Características Principales

01Automated quality checklist validation

02Structured XML specification generation

03Requirement-to-source traceability mapping

04Iterative clarification workflow (2-3 questions per cycle)

05Git-integrated documentation commits

060 GitHub stars

Casos de Uso

01Establishing a rigorous single source of truth for complex AI-driven codebases

02Validating functional requirements against formal acceptance criteria

03Converting vague project design documents into actionable technical specs

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add gregoryho/arcforge arc-refining

For use in Claude.ai and ChatGPT

Download Skill