create-epics

Generate production epics per architectural module from approved GDDs and ADRs.

1|Updated May 15, 2026
One-click install
npx skills add https://github.com/dnkrow/skill --skill create-epics-dnkrow
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: create-epics
Source: https://github.com/dnkrow/skill/tree/main/claude-global/create-epics
Command: npx skills add https://github.com/dnkrow/skill --skill create-epics-dnkrow

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill turns approved GDDs and architecture decisions into a set of clearly scoped epics per architectural module, so teams can align on ownership and boundaries before implementation starts.

Core Features & Use Cases

  • Epic-per-module generation: Creates one epic per architectural module and captures scope, layer, and ownership.
  • ADR coverage validation: Identifies governing ADRs, computes engine risk from the highest-risk ADR, and warns when requirements are untraced.
  • Controlled workflow with gates: Runs layer/system processing order, prompts for per-epic approval, and uses a producer assessment gate before writing files.
  • Next-step orchestration: Outputs epics and instructs you to run story creation for each epic afterward.

Quick Start

Run /create-epics layer: core to generate core-layer epics for each in-scope system.

Frequently Asked Questions about create-epics

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

FAQPage Schema
How do I generate epics from approved architecture and GDD documents?

Generating production epics from approved architecture involves translating GDDs and module ownership into one epic per architectural module, capturing scope, governing ADRs, engine risk, and requirement traceability before writing EPIC.md files.

What is requirements traceability in project planning for architectural modules?

Requirements traceability in project planning links approved GDD requirements to specific architectural modules and governing ADRs, ensuring every requirement is checked against accepted ADR coverage to identify untraced items and compute engine risk.

How do I create epics for a specific system or layer subset?

Creating epics for a specific layer or system subset requires running a targeted command like specifying the layer to process, generating only the in-scope architectural modules while maintaining ADR governance and requirement traceability checks.

Does this approach to epic generation require ADR coverage before writing files?

ADR coverage validation is required before writing files, as the workflow computes engine risk from the highest-risk ADR, warns when requirements are untraced, and enforces a producer assessment gate prior to creating production epic directories.

Can I stop at epic generation without creating individual stories?

Stopping at the epic level without creating stories is fully supported, as the workflow outputs epics and an index file after gate approvals, explicitly instructing you to run story creation for each epic afterward as a separate next step.

What happens when requirements are untraced against accepted ADRs during epic creation?

When requirements are untraced against accepted ADRs during epic creation, the system issues warnings identifying the gaps, calculates engine risk based on the highest-risk governing ADR, and prompts for explicit user approval before proceeding.