Product-Planning Skill
Analyzes product specifications for completeness, identifies gaps, and guides refinement to ensure requirements are ready for development.
What It Provides
Product Specification Analysis- Methods for reading flat and hierarchical product structures
- Patterns for identifying missing requirements
- Techniques for detecting ambiguous language
- Procedures for finding conflicting requirements
- Gap severity classification (blocking, warning, recommendation)
- Prioritization patterns for issue resolution
- Quality dimension evaluation methods
- Question formulation strategies (5 W's framework)
- Feature breakdown techniques
- Acceptance criteria templates
- Dependency mapping methods
- Domain organization verification
- Feature definition standards
- Priority level validation (CRITICAL/HIGH/MEDIUM/LOW)
Used By
/agentful-product Command:
- Primary consumer for product spec analysis
- Generates readiness scores and improvement recommendations
- Identifies gaps before development begins
- Uses this skill for comprehensive spec analysis
- Calculates readiness scores (0-100%)
- Validates tech stack compatibility
- Detects circular dependencies
- When user asks "Is my product spec ready?"
- When user uploads or updates product documentation
Tools Available
- Read (product specs, analysis reports)
- Write (analysis reports, product-analysis.json)
- Edit (product spec improvements)
- Glob (find domain files)
- Grep (search for features, requirements)
Product Structure Support
Flat Structure
.claude/product/
└── index.md # All features in one fileHierarchical Structure
.claude/product/
├── index.md # Overview
└── domains/
├── authentication/
│ └── index.md
└── profiles/
└── index.mdReadiness Scoring
IMPORTANT: The authoritative readiness scoring formula is defined in .claude/agents/product-analyzer.md.
- Readiness score ≥ 75%
- Zero blocking issues
- Tech stack 100% specified
- Completeness (features have descriptions, acceptance criteria)
- Clarity (unambiguous requirements, measurable metrics)
- Feasibility (realistic tech choices, achievable scope)
- Testability (verifiable acceptance criteria)
- Consistency (no conflicts, aligned priorities)
- BLOCKING: Cannot proceed without this information
- WARNING: Major ambiguity that will cause rework
- RECOMMENDATION: Nice-to-have details for better clarity
Analysis Workflow
Step 1: Detect Product Structure
# Check for flat structure
.claude/product/index.md
# Check for hierarchical structure
.claude/product/domains/*/index.mdStep 2: Analyze Completeness
For each feature, verify:- Description exists and is clear (what, why, who)
- Acceptance criteria are defined and testable
- Priority is assigned (CRITICAL/HIGH/MEDIUM/LOW)
- Technical considerations noted (performance, security, scalability)
- Tech stack is specified
- Goals are clear and measurable
- Target users are defined
- Domains are well-organized (hierarchical only)
Step 3: Delegate to Product Analyzer
For comprehensive analysis including readiness scoring:
Task("product-analyzer", "Analyze .claude/product/index.md and generate product-analysis.json")The product-analyzer will:
- Calculate readiness score (0-100%)
- Identify blocking issues, warnings, recommendations
- Validate priority levels
- Check tech stack compatibility
- Detect circular dependencies
Step 4: Generate Report
Save analysis to .agentful/product-analysis.json.
Step 5: Provide Recommendations
Generate specific, actionable recommendations:
- List specific questions to ask the user
- Suggest template improvements
- Recommend splitting/merging features
- Provide examples from similar features
Priority Level Guidelines
Standard priority levels (ONLY these are valid):
- CRITICAL: Must-have for MVP, blocking for launch
- HIGH: Important for launch, significant value
- MEDIUM: Nice-to-have, enhances experience
- LOW: Future consideration, not needed for launch
Invalid priority levels (flag as BLOCKING issue):
- P0, P1, P2, P3
- Must-have, Should-have, Nice-to-have
- 1, 2, 3, 4
- Any custom values
Examples
Feature Request Analysis
Feature: User Authentication
Status: Incomplete
Missing:
- OAuth provider support (should we support Google, GitHub?)
- Session timeout duration (security requirement)
- 2FA requirement (compliance consideration)
Questions:
1. Should we support OAuth providers (Google, GitHub)?
2. What's the session timeout requirement?
3. Do we need 2FA for compliance?Readiness Check
Product Specification Analysis
Readiness Score: 68/100
Ready to Build: NO
- Readiness score: 68% (>= 75% required)
- Blocking issues: 3 (0 required)
- Tech stack: 85% complete (100% required)
Blocking Issues (3):
1. **Authentication**: OAuth providers not specified
2. **Database**: Missing database choice
3. **Tech Stack**: Deployment platform not specifiedRefinement Guidance
Suggestions for "User Profiles" feature:
Current:
> Users can update their profiles
Improved:
> Users can update their profiles with:
> - Display name (required)
> - Avatar image (optional, max 2MB)
> - Bio text (optional, max 500 chars)
>
> Acceptance Criteria:
> - [x] User can update display name
> - [x] User can upload avatar (JPG, PNG)
> - [x] Avatar is resized to 200x200px
> - [x] Bio is limited to 500 characters
> - [x] Changes are saved immediatelyOutput Format
## Product Specification Analysis
### Readiness Score: X/100
[Brief summary of overall readiness]
### Ready to Build: YES/NO
- Readiness score: X% (>= 75% required)
- Blocking issues: X (0 required)
- Tech stack: X% complete (100% required)
### Blocking Issues (X)
- **[Feature/Domain]**: [Specific issue]
- **Why it's blocking**: [Explanation]
- **To resolve**: [Specific questions or required information]
### Warnings (X)
- **[Feature/Domain]**: [Specific gap]
- **Impact**: [What could go wrong]
- **Recommendation**: [How to address it]
### Recommendations (X)
- **[Feature/Domain]**: [Improvement area]
- **Benefit**: [Why this helps]
- **Suggestion**: [How to improve]
### Strengths
- [What's well-defined]
- [What's clear and actionable]
### Next Steps
1. [Most critical action]
2. [Second priority action]
3. [Third priority action]
---
*Analysis saved to `.agentful/product-analysis.json`*Rules
DO
- Use the product-analyzer agent for deep analysis
- Ask specific, targeted questions with concrete examples
- Save analysis results to
.agentful/product-analysis.json - Provide actionable recommendations
- Recognize well-written specifications
- Adapt analysis based on project type
- Consider technical feasibility
- Identify dependencies between features
- Validate priority levels are CRITICAL/HIGH/MEDIUM/LOW
DON'T
- Write code (this skill focuses on requirements analysis)
- Make assumptions about unclear requirements
- Provide generic feedback without specific questions
- Skip saving the analysis report
- Approve incomplete specs
- Ignore conflicting requirements
- Accept P0/P1/P2/P3 priority levels
- Suggest third-party services by default
Implementation Details
Location: .claude/skills/product-planning/SKILL.md
Model: Sonnet
Tools Used:- Read (product specs, analysis reports)
- Write (product-analysis.json)
- Edit (product spec improvements)
- Glob (find domain files)
- Grep (search for features)
.claude/product/index.md- Product specification.claude/product/domains/*/index.md- Domain specs.agentful/product-analysis.json- Analysis results
See Also
- /agentful-product - Uses this skill for product spec analysis
- Product-Analyzer Agent - Performs comprehensive analysis
- Product-Tracking Skill - Progress calculation methods