Don Cheli: PRD Generator

Instructions

1. Discover sources — Identify all available inputs:

• Figma links: analyze screens, flows, components, missing states
• Brief documents: extract objectives, audience, problems
• User research: identify personas, JTBD, pain points
• Existing code: detect technical constraints
• Conversation context: extract requirements mentioned by the user

1. Analyze Figma (if provided):

• Count screens/frames and map navigation flow
• Extract texts, labels, form fields, interactions
• Detect missing UI states: loading, empty, error, success, offline
• Identify gaps: flows not designed, edge cases missing
• Convert each screen → user stories → acceptance criteria

1. Risk analysis — For each feature, evaluate 7 risk categories:

• Technical (complexity, integrations, tech debt)
• Product (market fit, adoption, competition)
• UX (usability, accessibility, learning curve)
• Business (revenue, cost, time-to-market)
• Legal (GDPR, compliance, ToS)
• Security (OWASP, sensitive data, auth)
• Operational (support, scalability, monitoring)
• Assign probability (Low/Medium/High) and impact (Low/Medium/High/Critical)
• Propose mitigation for each risk

1. Prioritize features using MoSCoW + RICE:

• MoSCoW: Must / Should / Could / Won't
• RICE: (Reach × Impact × Confidence) / Effort
• Rank features by RICE score

1. Generate PRD with 13 sections:
2. Executive Summary
3. Problem (context, pain points, evidence)
4. Objectives & Success Metrics (KPIs, North Star)
5. Target Audience (personas, JTBD, user journey)
6. Proposed Solution (overview, main flow, features by priority)
7. Detailed Requirements (functional with Gherkin, non-functional, integrations, data model)
8. Design & UX (Figma analysis, UI states, responsive, accessibility)
9. Risk Analysis (matrix, mitigations, dependencies)
10. Scope (in scope, out of scope, future iterations)
11. Timeline & Milestones
12. Competitive Analysis
13. Launch Plan (rollout, monitoring, rollback)
14. Appendices (glossary, references, change history)

1. Validate PRD quality:

• Every requirement must be verifiable/testable
• Success metrics must be defined with baseline and target
• Every risk must have a mitigation
• Scope must be realistic for the timeline
• Anti-scope (out of scope) must be explicit

Output Format

markdown
# PRD: [Product/Feature Name]

## Metadata
| Field | Value |
|-------|-------|
| Version | 1.0 |
| Author | [name] |
| Status | Draft / In Review / Approved |
| Date | [date] |

## 1. Executive Summary
...
(continue with all 13 sections)

Save output to:

• .dc/prd/prd-v1.0.md — Full PRD
• .dc/prd/prd-risks.md — Expanded risk matrix
• .dc/prd/prd-figma-analysis.md — Figma analysis (if applicable)

Integration with SDD Pipeline

After generating the PRD, suggest:

/dc:specify    → Convert user stories into Gherkin specs
/dc:tech-plan  → Generate blueprint from the PRD
/dc:breakdown  → Create TDD tasks from the stories

Templates

Select based on product type:

• saas — B2B platforms, dashboards, multi-tenant
• mobile — Native iOS/Android, PWA
• ecommerce — Online stores, marketplaces
• api — Public APIs, SDKs, developer platforms
• internal — Internal team tools
• landing — Landing pages, conversion funnels

Quality Gate

• Never generate a PRD without at least 1 verifiable success metric
• Never omit the risk section — if no risks are visible, dig deeper
• Never leave scope ambiguous — mark unclear items as [NEEDS CLARIFICATION]
• Always include explicit anti-scope (what will NOT be built)
• Always generate user stories in Gherkin format for SDD integration
• Always version the PRD (v1.0, v1.1) with change history

Do not use this skill when

• The task is a simple bug fix or hotfix — use /dc:quick instead
• The user only needs technical specs — use /dc:specify directly
• The requirement is already well-defined in Gherkin — skip PRD, go to plan