PRD Tool¶
PRD Tool is a CLI and AI assistant integration for creating, validating, scoring, and viewing Product Requirements Documents (PRDs) using a canonical JSON schema.
Features¶
- Structured PRDs: JSON-based schema for consistent, machine-readable PRDs
- Quality Scoring: Automated rubric-based scoring with actionable recommendations
- Multiple Views: PM-focused and Executive summary views
- AI Integration: Works with Claude Code (MCP) and Kiro IDE (Power)
- Validation: Schema validation, ID format checking, and traceability verification
Quick Start¶
Install¶
Or build from source:
git clone https://github.com/agentplexus/agent-team-prd.git
cd agent-team-prd
go build -o bin/prdtool ./cmd/prdtool
Create a PRD¶
# Initialize a new PRD
prdtool init --title "User Authentication" --owner "Jane PM"
# Add content
prdtool add problem --statement "Users cannot securely access their accounts"
prdtool add persona --name "Developer Dan" --role "Backend Developer"
prdtool add goal --statement "Provide secure, seamless authentication"
prdtool add req --description "Support OAuth 2.0 login" --priority must
# Validate and score
prdtool validate
prdtool score
Use with AI Assistants¶
Documentation¶
- Installation - Detailed installation instructions
- CLI Reference - Complete command documentation
- CLI Examples - Common workflows and recipes
- Claude Code Integration - MCP server setup
- Kiro IDE Integration - Power installation
- PRD Schema Reference - JSON schema documentation
Scoring Categories¶
PRDs are scored across 10 categories:
| Category | Weight | Description |
|---|---|---|
| Problem Definition | 20% | Clear problem statement with evidence |
| Solution Fit | 15% | Solution addresses the problem |
| User Understanding | 10% | Personas and user journeys |
| Market Awareness | 10% | Competitive analysis |
| Scope Discipline | 10% | Clear goals and non-goals |
| Requirements Quality | 10% | Well-defined functional requirements |
| Metrics Quality | 10% | Measurable success criteria |
| UX Coverage | 5% | User experience considerations |
| Technical Feasibility | 5% | Technical constraints addressed |
| Risk Management | 5% | Risks identified with mitigations |
Score Thresholds:
- ≥8.0 → Approve (ready for implementation)
- ≥6.5 → Revise (minor issues to address)
- <6.5 → Human Review (significant gaps)
- ≤3.0 → Blocker (critical issues)