PRD Analyzer & Implementation Planner
Transform product requirements into actionable implementation plans with visual kanban tracking.
When NOT to Use
- Revising an existing plan — Use
plan-revision for modifying task hierarchies or scope changes - Starting implementation — Use
work-session to begin executing tasks from a plan - Quick one-off tasks — Use
/quick for tasks that don't need epic/story/task hierarchy - Investigating feasibility — Use
spike for time-boxed technical investigation before committing to a plan
Key Principles
- All outputs go to
.claude/ folder (PROJECT.md, tasks.db, features.json, kanban.html) - Break work into epic → story → task hierarchy (tasks ≤ 8h each)
- Assign skills to features based on type (api-design, database-design, etc.)
- Flag ambiguities before planning — don't assume unclear requirements
Quick Start Checklist
- Read the full PRD/brief document
- Extract goals, features, constraints, and dependencies
- Classify priorities (P0-P3) and flag ambiguities
- Break features into epics → stories → tasks with estimates
- Generate
.claude/ outputs (PROJECT.md, tasks.db, features.json, kanban.html) - Identify first skill to run and hand off
Skill Assignment
| Feature Type | Primary Skill | Secondary Skills |
|---|
| REST/GraphQL APIs | api-design | testing-strategy |
| Database schemas | database-design | — |
| SDK/library | sdk-development | — |
| Auth/Security | api-design | — |
| Testing infrastructure | testing-strategy | — |
| AWS/Cloud provisioning | cloud-infrastructure | — |
| Serverless/Lambda | cloud-infrastructure | api-design |
| Container deployment | cloud-infrastructure | ci-cd |
References
