gds-create-story

Generates comprehensive developer-ready story files from epics, architecture, and sprint status.

Updated Sep 3, 2026
One-click install
npx skills add https://github.com/watchthelight/shatterfish --skill gds-create-story-watchthelight
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: gds-create-story
Source: https://github.com/watchthelight/shatterfish/tree/main/.claude/skills/gds-create-story
Command: npx skills add https://github.com/watchthelight/shatterfish --skill gds-create-story-watchthelight

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Writing implementation stories by hand often leaves developers without the architecture constraints, previous-story learnings, and project conventions they need, causing rework and regressions. This Skill automates the creation of a complete, context-rich story file so the dev agent has everything required for implementation. ## Core Features & Use Cases - Automatic story discovery: Reads sprint-status.yaml to find the next backlog story, or accepts an explicit epic-story identifier like "1-2-user-auth". - Exhaustive context gathering: Loads epics, GDD, architecture, UX docs, project-context.md, previous story files, and recent git history to extract requirements and guardrails. - Quality validation: Validates the generated story against a checklist that hunts for missing requirements, wrong libraries, and regression risks before marking it ready-for-dev. - Use Case: A game development team finishes sprint planning and says "create the next story"; the Skill auto-selects the next backlog story, assembles all technical context, writes the story file, and updates sprint status to ready-for-dev. ## Quick Start Ask the agent to create the next story from the sprint backlog, or specify a story such as "create story 2-4".

Frequently Asked Questions about gds-create-story

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

FAQPage Schema
How do I create the next story from my sprint backlog?

Run the create-story workflow with no arguments and it reads sprint-status.yaml, finds the first story with backlog status, and generates the full story file. You can also pass an explicit identifier like "1-2-user-auth" to target a specific story.

What documents does the create-story workflow need as input?

It loads epics, architecture, GDD, and UX markdown files from your planning artifacts directory, plus project-context.md if it exists. Missing files are tolerated; the workflow notes them as unavailable and continues with what it finds.

Can I customize the story creation workflow for my team?

Yes, the customize.toml file supports team and user override files that merge with defaults. You can add activation steps, persistent facts, and an on_complete hook without editing the base skill.

What happens if no backlog stories exist in sprint-status.yaml?

The workflow halts and reports that all stories are created, in progress, or done. It suggests running sprint-planning to refresh tracking or using correct-course to add more stories.

How does the skill prevent developer implementation mistakes?

It extracts architecture constraints, reads files being modified, analyzes previous story learnings and git history, then validates the output against a checklist covering wrong libraries, regressions, and vague requirements before marking the story ready-for-dev.