hongkongball/myskills/build-prd
Claude Code and agent skill by hongkongball/myskills
- community
Checking that the SKILL.md file is still in the repository.
Convert the scoped hackathon idea into user-facing product requirements.
Install
npx vskill@latest install hongkongball/myskills/build-prdSKILL.md
The same SKILL.md was published earlier as hongkongball/myskills/build-prd.
Guided Build: PRD
Read references/build-guide.md, then follow this command.
This is the Codex version of the learning curriculum's PRD command.
Goal
Turn scope.md into a product requirements document. This step is about user behavior and acceptance criteria, not code structure.
Preconditions
Read .devpost-hackathon-state.json.
If the state file does not exist, direct the user to $start-hackathon.
Read everything in docs/hackathon-build/. If scope.md does not exist, direct the user to $build-scope.
Flow
Interview in small batches of related questions (per the build guide). You are a sharp interviewer here — no code talk, no technical decisions; pure "what does this thing need to do?" If implementation questions come up, defer warmly: "Great question — we'll get into that in $build-spec. For now, what does the user experience?"
Mandatory beats:
- Walk the scope section by section, converting brainstorm language into precise
behavior: "You said the app helps people find X. What does a user see when they first open it?" Zoom in relentlessly (Sharpening Questions in the build guide).
- Organize behaviors into user stories and epics with stable headings — introduced
without jargon: "Let me capture what you're describing — 'As a [person], I want [thing] so that [reason].' Does that match?"
- Testable acceptance criteria per story: "How would you know this is working? What would
you see on screen?" Not vague ("search works well"), not implementation ("query under 100ms"), not untestable ("the UX is intuitive").
- Edge cases: empty states, first-run experience, error cases, "what if they do X before
Y?" Aim for 2-3 genuine "oh, I hadn't thought of that" moments.
- Guard scope against the time budget recorded in scope.md. When something grows: "This
is getting bigger than the time you have. Essential for your submission, or would you add it later?" That one question sorts everything into What We're Building versus What We'd Add With More Time. Keep Non-Goals specific and reasoned — "NOT building user profiles, because the app works fine anonymous" beats "no extra features."
Occasionally make the expansion visible: "See how much more specific we're getting? The scope said 'users can search' — now we know exactly what that means."
After mandatory beats, offer a deepening round per the build guide. Good PRD deepening topics: feature interactions ("if a user changes X while looking at Y, what should happen?"), persistence ("close the app and come back — is their stuff still there?"), boundary cases ("what if someone adds 100 of these?"), the Devpost "wow moment" ("which feature makes someone stop scrolling on the submission page?"), user-order assumptions ("you're assuming they do X first — what if they don't?"), and what would make it feel really good, not just functional.
Output
Use references/templates/prd-template.md.
Create or update:
docs/hackathon-build/prd.mddocs/hackathon-build/build-notes.md
The PRD should feel significantly more substantial than the scope doc. If it's roughly the same length, you haven't expanded enough.
State Update
Set:
learning.current_steptoprd- add
scopetolearning.completed_stepsif missing next_commandtobuild-spec
Presentation Output
Compose the response in-context per references/plugin-runtime.md ("Composing the Response"): read references/content/learning/prd.md, strip maintainer <!-- --> comments, interpolate the event name, then present a short stage headline, the page content, and the next-step callout. Do not run any script. End by recommending $build-spec.
Required References
references/plugin-runtime.mdreferences/build-guide.mdreferences/content/learning/prd.md
Security
- Security scanPassed, 90 of 100Pattern checks on the skill's files.
- Model reviewPassed
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingskills/plugins/openai-curated-remote/app-6a330a7730c081919892632d5baaec58/4.0.1/skills/build-prd/SKILL.md
Popularity
- 7-day trend score
- 0
Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/hongkongball/myskills/build-prd)