lhstorm/mcp_server_prompt_library

MCP Prompt Library Server

Professional-grade AI prompts for domain experts. This reusable MCP server delivers expert-level prompt libraries that transform how you work with Claude. Currently featuring supercharged Product Management prompts, but easily adaptable to any domain.

🎯 What You Get

Transform your Claude experience with expert-level prompts that provide:

• Structured methodologies (RICE, INVEST, SCQA frameworks)
• Comprehensive deliverables (7-12 outputs per prompt)
• Real-world examples with specific metrics and contexts
• Professional best practices from senior practitioners

Example: Instead of basic user story writing, get stories with INVEST principles, Given-When-Then acceptance criteria, edge cases, non-functional requirements, and technical integration points.

🚀 Quick Start

⚡ Option 1: Use Product Management Prompts (Ready Now)

git clone <this-repository>
cd mcp-prompt-library
python -m venv venv
source venv/bin/activate  # Windows: venv\Scripts\activate
pip install -r requirements.txt
python tools/configure-claude.py

Restart Claude Desktop and you're ready!

🛠️ Option 2: Create Your Domain

python setup.py

Interactive wizard configures any domain in 2 minutes.

📋 How to Use in Claude Desktop

Once configured, interact naturally with Claude:

🔍 Discovery Commands:

• "What prompts do I have available?"
• "List all strategy prompts"
• "Search for prompts about user research"

📝 Using Prompts:

• "Use the user-story-writer prompt for a mobile checkout feature"
• "Get the feature-prioritization prompt and help me prioritize my Q2 roadmap"
• "Apply the competitive-analysis prompt to analyze project management tools"

🎯 Real Usage Examples:

You: "Use the user-story-writer prompt for a password reset feature"

Claude: [Loads expert prompt] I'll help you create a comprehensive user story using INVEST principles and Given-When-Then acceptance criteria...

You: "Search for analytics prompts"

Claude: I found 1 analytics prompt:
- Product Metrics & KPI Analyst: Defines measurement frameworks with North Star metrics...

You: "Use that analytics prompt for our new onboarding flow"

Claude: [Loads comprehensive analytics framework] I'll design a measurement framework including leading indicators, cohort analysis, and A/B testing plans...

🏆 Featured: Supercharged Product Management Prompts

1. User Story Writer

Expert-Level Features:

• INVEST principles validation
• Given-When-Then acceptance criteria
• Accessibility, performance, security considerations
• Dependency mapping and integration points

Example Output:

User Story: Password Reset with MFA
As Sarah (returning customer, 35, mobile-first, moderate tech skills)
I want to reset my password securely without calling support
So that I can quickly regain access and continue my premium subscription

Acceptance Criteria:
Given I'm on the login screen
When I tap "Forgot Password" 
Then I should see options for email or SMS verification...

[Continues with 7 comprehensive sections]

2. Feature Prioritization Framework

Methodologies Included:

• Detailed RICE scoring with calculations
• Value vs Effort matrix positioning
• Kano model categorization
• Strategic alignment assessment
• Risk analysis and mitigation

Sample Input:

Features: Dark mode, Advanced search, Push notifications
Constraints: 2 developers, 3-month timeline, $150K budget
Objectives: Increase DAU by 25%, reduce churn by 15%
User base: 45,000 MAU, 60% mobile

3. Competitive Analysis Expert

Analysis Framework:

• Multi-dimensional competitive mapping
• Jobs-to-be-Done framework analysis
• Pricing strategy deconstruction
• Market share and growth analysis
• Whitespace opportunity identification

4. Product Metrics & KPI Analyst

Measurement Framework:

• North Star Metric identification
• Leading vs lagging indicator mapping
• A/B testing framework design
• Cohort retention analysis
• Real-time dashboard specifications

5. Customer Interview Guide Creator

Research Methodologies:

• Behavioral interview techniques
• Cognitive bias mitigation strategies
• Progressive disclosure questioning
• Cultural sensitivity guidelines
• Post-interview analysis framework

6. PRD Writer

10-Section Framework:

• Executive summary with business case
• User research insights and validation
• Technical architecture specifications
• API integration requirements
• Risk analysis and contingency planning

7. Go-to-Market Strategist

12-Point GTM Framework:

• Market segmentation and TAM analysis
• Customer segmentation and ICP definition
• Pricing strategy with sensitivity analysis
• Channel strategy and partnership development
• Launch timeline with milestone tracking

8. Stakeholder Communication Expert

Communication Framework:

• SCQA structure (Situation, Complication, Question, Answer)
• Audience analysis and stakeholder mapping
• Executive presence development
• Q&A preparation and objection handling

🎨 Customization Options

Switch to Different Domains

# Use Software Engineering prompts
cp templates/software-engineering.json prompts.json
python tools/configure-claude.py

# Use Marketing prompts  
cp templates/marketing.json prompts.json
python tools/configure-claude.py

Create Your Own Domain

1. AI-Generated (Recommended):

   • Open docs/prompt-generator-template.md
   • Customize for your domain (Legal, Healthcare, Sales, etc.)
   • Send to Claude: "Generate prompts for [your domain] using this template"
   • Copy JSON response to prompts.json

2. Manual Creation:

   {
     "id": "unique-id",
     "name": "Expert Name", 
     "description": "What this expert does",
     "category": "category",
     "tags": ["tag1", "tag2"],
     "system": "You are a world-class expert with...",
     "template": "Comprehensive template with {variables}",
     "example_usage": {
       "variable": "realistic example"
     }
   }
   

Server Configuration

Edit server-config.json:

{
  "server_name": "legal-prompts",
  "display_name": "Legal Expert Prompts",
  "description": "AI prompts for legal professionals", 
  "domain": "legal",
  "categories": ["contracts", "research", "compliance"]
}

Note: Use python tools/configure-claude.py to automatically configure Claude Desktop. If you need manual configuration, see claude-config.example.json for the format.

🔧 Professional Tools

# Validate prompt quality and structure
python tools/validate-prompts.py

# Auto-configure Claude Desktop (handles paths automatically)
python tools/configure-claude.py

# Preview configuration without changes
python tools/configure-claude.py --dry-run

# Check current server settings
python tools/configure-claude.py --show-config

📁 Project Architecture

mcp-prompt-library/
├── 📄 server.py                    # MCP server (configurable)
├── ⚙️ server-config.json           # Domain configuration
├── 🎯 prompts.json                # Active expert prompts
├── 🚀 setup.py                    # Interactive domain setup
├── 📋 claude-config.example.json   # Claude Desktop config example
├── 🛠️ tools/                      # Professional utilities
│   ├── configure-claude.py        # Auto-configuration
│   └── validate-prompts.py        # Quality validation
├── 📚 templates/                  # Expert domain libraries
│   ├── product-management.json    # 8 supercharged PM prompts
│   ├── software-engineering.json  # 6 engineering prompts
│   └── marketing.json             # 6 marketing prompts
└── 📖 docs/                       # Comprehensive guides
    ├── setup-guide.md
    └── prompt-generator-template.md

🌟 Why These Prompts Are Different

❌ Basic Prompts Give You:

• Generic advice
• Simple templates
• Basic examples
• Single methodology

✅ Our Expert Prompts Deliver:

• Multiple methodologies (RICE + Kano + Value vs Effort)
• Comprehensive frameworks (10-section PRDs, 12-point GTM strategies)
• Realistic examples ("Sarah, 35, mobile-first user" vs "a user")
• Professional rigor (Given-When-Then criteria, risk assessments)
• Actionable deliverables (7-12 specific outputs per prompt)

📊 Prompt Quality Comparison:

🎯 Use Cases & Workflows

Product Managers:

1. Morning standup prep: "Use metrics analyst for our retention dashboard review"
2. Feature planning: "Apply prioritization framework to these 8 features"
3. Stakeholder meetings: "Create executive communication for roadmap changes"

Consultants:

1. Switch domains per client: cp templates/marketing.json prompts.json
2. Generate client-specific prompts using AI template
3. Validate quality: python tools/validate-prompts.py

Teams:

1. Shared expertise: Everyone gets senior-level guidance
2. Consistent quality: Standardized methodologies across team
3. Rapid onboarding: Junior members learn from expert prompts

🚨 Troubleshooting

Server Connection Issues

# Check server health
python -c "from server import load_prompts; print(f'✅ {len(load_prompts())} prompts loaded')"

# Validate configuration  
python -c "from server import SERVER_CONFIG; print(SERVER_CONFIG)"

# Test manually
python server.py

Claude Desktop Not Finding Server

1. Verify config location: ~/Library/Application Support/Claude/claude_desktop_config.json
2. Use absolute paths (our tool handles this automatically)
3. Restart Claude Desktop completely
4. Check logs: Help → View Logs in Claude Desktop

Prompt Validation Errors

# Check prompt structure
python tools/validate-prompts.py --strict

# Common fixes:
# - Use lowercase, hyphenated IDs
# - Include {placeholder} variables in templates  
# - Provide realistic example_usage values

Quick Health Checks

# Test configuration loading
python tools/configure-claude.py --show-config

# Validate all prompts
python tools/validate-prompts.py

# Preview Claude config changes
python tools/configure-claude.py --dry-run

🎮 Demo: See It In Action

Request: "Help me prioritize features for our mobile app"

Without Our Prompts:

"I can help you prioritize features. What features are you considering?"

With Our Expert Prompts:

"I'll use our comprehensive feature prioritization framework that combines RICE scoring, Value vs Effort analysis, and Kano model categorization. Please provide:

Features to evaluate: [Your feature list] Resource constraints: [Team size, timeline, budget]
Strategic objectives: [Specific goals with metrics] User base size: [MAU and segments] Available data: [Research, analytics, surveys] Competitive pressure: [Market context]

I'll deliver:

1. RICE scoring with detailed calculations
2. Value vs Effort matrix positioning
3. Kano model categorization
4. Strategic alignment assessment
5. Risk analysis and mitigation strategies
6. Final prioritized roadmap with rationale"

🚀 Getting Started Checklist

• Clone repository
• Set up virtual environment
• Run python tools/configure-claude.py
• Restart Claude Desktop
• Test: "What prompts do I have?"
• Try: "Use the user-story-writer prompt"
• Explore: Browse all 8 expert prompts
• Customize: Add your domain-specific prompts

🤝 Contributing & Community

Add New Domains:

1. Create templates/your-domain.json
2. Include 6-8 expert-level prompts
3. Test with validation tools
4. Submit PR with examples

Current Templates:

• ✅ Product Management (8 prompts) - Supercharged!
• ✅ Software Engineering (6 prompts)
• ✅ Marketing (6 prompts)
• 🔄 Coming Soon: Legal, Healthcare, Sales, Finance

📄 License

MIT License - Use freely for personal and commercial projects.

🚀 Ready to transform your Claude experience? Start with python tools/configure-claude.py and see the difference expert-level prompts make.