skills/prd-creator/SKILL.md
Transforms a vague project idea into a structured PRD through targeted conversation, then hands off to greenfield-init for full project planning. Use when a user has a rough app idea, wants to build something new, or needs to turn a concept into a documented scope before planning begins.
npx skillsauth add barkbarkgoose/ai-agents prd-creatorInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Turn a rough idea into a structured Product Requirements Document, then feed it directly into the greenfield planning pipeline.
user prompt → [prd-creator] → PRD.md → [greenfield-init] → PROJECT_BLUEPRINT.md → [greenfield-decomposer] → phase research docs
Accept the user's basic description. It can be one sentence or a paragraph — whatever they have.
Use AskQuestion to gather the essentials. Group related questions. Target 4–6 questions max:
After Round 1, ask a focused follow-up to define scope edges:
Use AskQuestion again if the answers will branch the PRD significantly. Otherwise, ask conversationally.
Using answers from both rounds, write a PRD following the template in PRD_TEMPLATE.md.
Show the user a summary of the PRD and ask for confirmation or corrections before saving.
Save the confirmed PRD to:
.agent-tasks/prd-drafts/[project-slug]-PRD.md
Where [project-slug] is a lowercase-hyphenated short name derived from the project title.
Create the directory if it doesn't exist.
After saving, instruct the user:
PRD saved. Next step: use the
greenfield-initskill to turn this into a project blueprint.Use the greenfield-init skill with PRD: .agent-tasks/prd-drafts/[project-slug]-PRD.md
The user can trigger this immediately, or you can invoke it directly if they confirm they're ready to proceed.
Before saving the PRD:
tools
Use this skill when working on Vue 3 + TypeScript client-side code, including creating new components, refactoring existing UI, implementing store logic with Pinia, or building reusable composition functions.
data-ai
orchestration skill for tasks, takes a task folder as input and runs one sub-agent for each individual task file. Should NOT execute or make any changes on its own, only sub-agents may do that.
tools
create tasks as files in local project directory
data-ai
archives a local agent task directory so it can be recalled for future reference