Prd patterns
Skill Mattakushi432/Claude-Code-Skills-Custom-DevTools-Pack/plugins/devtools-pack/skills/prd-patterns
A curated pack of custom Claude Code skills for developers — installable as a Claude Code plugin marketplace.
npx -y skills add Mattakushi432/Claude-Code-Skills-Custom-DevTools-Pack --skill prd-patternsAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
What its author says it does
Copied from the file, not written here
When to activate: PRD, product requirements document, feature spec, functional spec, requirements writing, acceptance criteria, edge cases
SKILL.md
5.8 KB, as published. Nobody here has run it
PRD Patterns
PRD Structure Template
# [Feature Name] PRD
**Status**: Draft | In Review | Approved | Shipped
**Author**: [Name]
**Reviewers**: [Engineering Lead], [Design Lead], [Data], [Legal if needed]
**Last Updated**: YYYY-MM-DD
**Target Release**: Q[X] YYYY
---
## 1. Problem Statement
One paragraph. What user pain exists? What evidence supports it?
## 2. Goals & Success Metrics
- Goal 1: [Outcome, not output]
- Success metric: [Specific measurable target with baseline]
## 3. Non-Goals
- Explicitly exclude scope to prevent creep
## 4. User Stories
[See user-stories.md for format]
## 5. Functional Requirements
FR-001: [Requirement]
FR-002: [Requirement]
## 6. Non-Functional Requirements
NFR-001: Latency < 200ms at p95
NFR-002: Availability 99.9%
## 7. UX/Design
[Figma link] [Key screens described]
## 8. Technical Considerations
[Architecture notes, dependencies, risks]
## 9. Edge Cases & Error States
[Table of edge cases]
## 10. Analytics Plan
[Events to instrument]
## 11. Launch Plan
[Phased rollout, feature flags, comms]
## 12. Open Questions
| Question | Owner | Due |
|----------|-------|-----|
Acceptance Criteria Writing
GIVEN / WHEN / THEN Format
GIVEN [context/precondition]
WHEN [action taken]
THEN [expected outcome]
AND [additional assertion]
Example:
GIVEN a user is logged in with a free plan
WHEN they click "Export to CSV"
THEN they see an upgrade modal
AND the modal shows Pro plan features
AND clicking "Upgrade" redirects to /billing
Checklist Format (simpler features)
AC-001: User can filter by date range
- [ ] Start date defaults to 30 days ago
- [ ] End date defaults to today
- [ ] Invalid range (start > end) shows inline error
- [ ] Maximum range: 365 days; exceeding shows error message
- [ ] Empty result set shows empty state with CTA
Edge Cases Documentation
Edge Case Table
| Scenario | Input | Expected Behavior | Priority |
|---|---|---|---|
| Empty state | No data returned | Show empty state UI with CTA | P0 |
| Network error | API timeout > 5s | Show retry toast, do not lose data | P0 |
| Concurrent edit | Two users edit same record | Last-write-wins with timestamp | P1 |
| Invalid characters | SQL injection attempt | Sanitize input, reject gracefully | P0 |
| Max payload | File > 10MB upload | Show size limit error before upload | P1 |
| Session expired | Token expires mid-flow | Redirect to login, preserve intent | P0 |
Error States Checklist
- 400 Bad Request: show user-friendly validation message
- 401 Unauthorized: redirect to login
- 403 Forbidden: show "you don't have access" message
- 404 Not Found: show 404 page with navigation
- 429 Rate Limited: show "too many requests, try again in X seconds"
- 500 Server Error: show generic error with support link
- Network offline: show offline banner, queue actions if feasible
Requirement Writing Rules
DO
- Use "the system SHALL" for mandatory requirements
- Use "the system SHOULD" for recommended
- Use "the system MAY" for optional
- Reference exact UI element names from design
- Include specific numbers (thresholds, limits, timeouts)
- Attach Figma links to specific frames
DON'T
- Write vague requirements: "the page should load fast" → "p95 < 2s"
- Combine multiple behaviors in one requirement
- Leave decisions open: "TBD" in approved PRDs is a blocker
- Describe implementation, only describe behavior
Stakeholder Sign-Off Process
Review Stages
1. Author Draft (Day 0)
→ Self-review against checklist below
2. Design Review (Day 1-2)
→ UX completeness, design feasibility
3. Engineering Review (Day 3-5)
→ Technical feasibility, estimates, NFRs
4. Data Review (Day 5-6)
→ Analytics plan completeness, metric definitions
5. Legal/Compliance (if needed, Day 5-7)
→ Privacy, regulatory requirements
6. PM/Leadership Approval (Day 7-10)
→ Priority alignment, resource commitment
7. Kickoff Meeting
→ All stakeholders aligned, questions resolved
PRD Self-Review Checklist
- Problem clearly stated with supporting evidence (qualitative + quantitative)
- Goals are outcomes, not outputs
- Success metrics have baseline and target values
- Non-goals explicitly listed
- All user roles/personas addressed
- Every functional requirement has at least one acceptance criterion
- All edge cases documented
- Error states defined
- NFRs specified (performance, availability, security)
- Analytics instrumentation plan included
- Dependencies on other teams identified
- Rollout plan defined (flag? % rollout? which segments?)
- Open questions resolved or assigned owners
PRD Status Definitions
| Status | Meaning |
|---|---|
Draft | Author working, not ready for review |
In Review | Circulating to stakeholders |
Revision Needed | Feedback received, author updating |
Approved | All required sign-offs obtained |
In Progress | Engineering actively building |
Shipped | Feature live in production |
Deprecated | Feature killed or superseded |
Minimal PRD (for small features)
For < 2 sprint features, use this lighter format:
## What
[1-2 sentence description]
## Why
[Problem + evidence]
## How (Behavior)
- [Bullet list of user-visible behaviors]
## Acceptance Criteria
- [ ] AC1
- [ ] AC2
## Not included
- [Explicit scope exclusions]
## Metrics
- Primary: [metric]
- Secondary: [metric]