DollhouseMCP
A comprehensive Model Context Protocol server that enables dynamic AI persona management with GitHub-powered marketplace integration, allowing Claude and other compatible AI assistants to activate different behavioral personas with community sharing capabilities.
README
DollhouseMCP
Platform Support
A comprehensive Model Context Protocol (MCP) server that enables dynamic AI persona management with an integrated GitHub-powered marketplace. DollhouseMCP allows Claude and other compatible AI assistants to activate different behavioral personas while supporting community sharing and monetization.
🌐 Repository: https://github.com/mickdarling/DollhouseMCP
🏪 Marketplace: https://github.com/mickdarling/DollhouseMCP-Personas
🌍 Website: https://dollhousemcp.com (planned)
✨ Key Features
| Feature | Description |
|---|---|
| 🎭 23 MCP Tools | Complete persona lifecycle management through chat interface |
| 🏪 GitHub Marketplace | Browse, search, install, and submit personas to community marketplace |
| 👤 User Identity System | Environment-based attribution for persona creators |
| 🆔 Unique ID System | Advanced ID generation: what-it-is_YYYYMMDD-HHMMSS_who-made-it |
| 💬 Chat-Based Management | Create, edit, and validate personas through conversational interface |
| 🔄 Real-time Operations | Live editing with automatic version bumping and validation |
| 🚀 Auto-Update System | Enterprise-grade auto-update with backup/rollback and dependency validation |
| 🏠 Local-First Architecture | Full functionality without cloud dependency |
🚀 Quick Start
Installation
Automated Setup (Recommended) - Claude Desktop Only
[!WARNING] Claude Desktop Only: The automated setup script is specifically designed for Claude Desktop integration. If you're using Claude Code, other AI platforms (ChatGPT, BoltAI, Gemini, etc.), or custom MCP implementations, please use the Manual Installation process below.
# Clone the repository
git clone https://github.com/mickdarling/DollhouseMCP.git
cd DollhouseMCP
# Run automated setup script (Claude Desktop only)
./setup.sh
The setup script will:
- 📦 Install all dependencies
- 🔨 Build the TypeScript code
- 📍 Detect your installation path
- 🔧 Generate the exact Claude Desktop configuration
- 📋 Provide step-by-step setup instructions
Manual Installation
Note: Manual installation works with all MCP-compatible platforms including Claude Desktop, Claude Code, ChatGPT, BoltAI, Gemini, and custom implementations.
# Clone the repository
git clone https://github.com/mickdarling/DollhouseMCP.git
cd DollhouseMCP
# Install dependencies and build
npm install
npm run build
# Optional: Set user identity for persona attribution
export DOLLHOUSE_USER="your-username"
export DOLLHOUSE_EMAIL="your-email@example.com"
Claude Desktop Configuration
Add DollhouseMCP to your Claude Desktop configuration:
Location: ~/Library/Application Support/Claude/claude_desktop_config.json (macOS)
{
"mcpServers": {
"dollhousemcp": {
"command": "node",
"args": ["/absolute/path/to/DollhouseMCP/dist/index.js"]
}
}
}
🔄 After configuration:
- Save the file
- Restart Claude Desktop completely
- All 23 DollhouseMCP tools will be available
🛠️ Available Tools (23 Total)
Core Persona Management
list_personas- Display all local personas with enhanced metadataactivate_persona- Activate by name, filename, or unique IDget_active_persona- Get current persona informationdeactivate_persona- Return to default modeget_persona_details- View complete persona detailsreload_personas- Refresh from filesystem
GitHub Marketplace Integration
browse_marketplace- Browse personas by categorysearch_marketplace- Search across all marketplace personasget_marketplace_persona- View detailed marketplace persona infoinstall_persona- One-click download and installationsubmit_persona- Submit to marketplace via GitHub issue
User Identity Management
set_user_identity- Set username for persona attributionget_user_identity- View current identity statusclear_user_identity- Return to anonymous mode
Chat-Based Persona Management
create_persona- Guided persona creation through chatedit_persona- Modify existing persona fieldsvalidate_persona- Comprehensive quality validation
Auto-Update System
check_for_updates- Check GitHub releases for available DollhouseMCP updatesupdate_server- Automated git pull + npm install + build with backup creationrollback_update- Restore previous version from automatic backupsget_server_status- Comprehensive server status with version, git info, and system details
Persona Indicators
configure_indicator- Configure how persona indicators appear in AI responsesget_indicator_config- View current indicator configuration settings
📖 Usage Examples
Marketplace Operations
browse_marketplace # See all categories
browse_marketplace "creative" # Browse creative personas
search_marketplace "writing" # Search for writing personas
install_persona "creative/storyteller.md" # Install from marketplace
Persona Creation & Management
create_persona "Study Buddy" "A helpful tutor" "educational" "You are a patient tutor..."
edit_persona "Study Buddy" "description" "An encouraging academic mentor"
validate_persona "Study Buddy" # Check quality and format
submit_persona "Study Buddy" # Share with community
User Identity
set_user_identity "your-username" # Set attribution
get_user_identity # Check current status
clear_user_identity # Return to anonymous
Auto-Update Operations
check_for_updates # Check for new DollhouseMCP versions
get_server_status # View current version and system info
update_server true # Perform automated update with backup
rollback_update true # Revert to previous version if needed
Persona Indicators
DollhouseMCP adds visual indicators to AI responses when a persona is active:
[🎭 Creative Writer v2.1 by @mickdarling] Your creative response here...
Configure indicators:
get_indicator_config # View current settings
configure_indicator enabled:false # Disable indicators
configure_indicator style:"minimal" # Use minimal format: 🎭 Creative Writer
configure_indicator style:"compact" # Use compact: [Creative Writer v2.1]
configure_indicator style:"full" # Full format (default)
configure_indicator emoji:"🤖" # Change emoji
configure_indicator showAuthor:false # Hide author attribution
configure_indicator bracketStyle:"round" # Use (parentheses) instead of [brackets]
Environment variables for persistent configuration:
export DOLLHOUSE_INDICATOR_ENABLED=true
export DOLLHOUSE_INDICATOR_STYLE=minimal
export DOLLHOUSE_INDICATOR_EMOJI=🎨
🖥️ Cross-Platform Installation
🐧 Linux Installation
Prerequisites
# Ubuntu/Debian
sudo apt update
sudo apt install -y nodejs npm git
# CentOS/RHEL/Fedora
sudo dnf install -y nodejs npm git
# Arch Linux
sudo pacman -S nodejs npm git
Installation Steps
# Clone and build
git clone https://github.com/mickdarling/DollhouseMCP.git
cd DollhouseMCP
npm install
npm run build
# Optional: Set user identity
export DOLLHOUSE_USER="your-username"
export DOLLHOUSE_EMAIL="your-email@example.com"
Claude Desktop Configuration (Linux)
# Configuration location
~/.config/Claude/claude_desktop_config.json
# Or use XDG_CONFIG_HOME if set
$XDG_CONFIG_HOME/Claude/claude_desktop_config.json
Configuration content:
{
"mcpServers": {
"dollhousemcp": {
"command": "node",
"args": ["/absolute/path/to/DollhouseMCP/dist/index.js"]
}
}
}
🪟 Windows Installation
Prerequisites
# Install Node.js from https://nodejs.org/
# Or using Chocolatey
choco install nodejs git
# Or using winget
winget install OpenJS.NodeJS Git.Git
Installation Steps (PowerShell)
# Clone and build
git clone https://github.com/mickdarling/DollhouseMCP.git
cd DollhouseMCP
npm install
npm run build
# Optional: Set user identity
$env:DOLLHOUSE_USER = "your-username"
$env:DOLLHOUSE_EMAIL = "your-email@example.com"
Claude Desktop Configuration (Windows)
# Configuration location
$env:APPDATA\Claude\claude_desktop_config.json
Configuration content (use forward slashes or double backslashes):
{
"mcpServers": {
"dollhousemcp": {
"command": "node",
"args": ["C:/path/to/DollhouseMCP/dist/index.js"]
}
}
}
🍎 macOS Installation
Prerequisites
# Using Homebrew (recommended)
brew install node git
# Or download from https://nodejs.org/
Installation Steps
# Clone and build
git clone https://github.com/mickdarling/DollhouseMCP.git
cd DollhouseMCP
npm install
npm run build
# Optional: Set user identity
export DOLLHOUSE_USER="your-username"
export DOLLHOUSE_EMAIL="your-email@example.com"
Claude Desktop Configuration (macOS)
# Configuration location
~/Library/Application Support/Claude/claude_desktop_config.json
Configuration content:
{
"mcpServers": {
"dollhousemcp": {
"command": "node",
"args": ["/absolute/path/to/DollhouseMCP/dist/index.js"]
}
}
}
🐳 Docker Installation
Quick Start with Docker
# Clone repository
git clone https://github.com/mickdarling/DollhouseMCP.git
cd DollhouseMCP
# Build and run with Docker Compose
docker-compose up -d
# Or build manually
docker build -t dollhousemcp .
docker run -d --name dollhousemcp dollhousemcp
Docker Compose (Recommended)
Production deployment:
docker-compose up -d
Development with hot reload:
docker-compose --profile dev up dollhousemcp-dev
Custom Personas with Docker
# Mount your custom personas directory
docker run -d \
--name dollhousemcp \
-v /path/to/your/personas:/app/personas \
-e DOLLHOUSE_USER="your-username" \
dollhousemcp
Docker Environment Variables
# Set user identity
DOLLHOUSE_USER=your-username
DOLLHOUSE_EMAIL=your-email@example.com
# Custom personas directory (inside container)
PERSONAS_DIR=/app/personas
# Node.js environment
NODE_ENV=production
🧪 Testing
Running Tests
The project includes comprehensive tests for cross-platform compatibility:
# Run all tests
npm test
# Run tests with coverage
npm run test:coverage
# Run tests in watch mode (for development)
npm run test:watch
# Run specific test suites
npm run test:auto-update
npm run test:personas
npm run test:marketplace
Test Coverage
Current test coverage includes:
- ✅ 102 comprehensive tests covering all functionality
- ✅ Auto-update system - GitHub API, backup/rollback, dependency validation
- ✅ Security testing - Command injection prevention, input validation
- ✅ Cross-platform compatibility - Windows, macOS, Linux path handling
- ✅ Version validation - Parsing tests for git/npm output formats
- ✅ Edge case coverage - Network failures, missing dependencies, malformed input
Manual Verification
Verify your setup works correctly:
# Build the project
npm run build
# Test the server (should output server info)
node dist/index.js --help 2>/dev/null || echo "Server compiled successfully"
# Verify personas directory
ls -la personas/
# Test auto-update system
check_for_updates # Use in Claude Desktop
get_server_status # Check current system status
☁️ Cloud Deployment
Container Registries
The project supports deployment to:
- GitHub Container Registry (ghcr.io)
- Docker Hub
- AWS ECR
- Google Container Registry
Example Cloud Deployments
AWS ECS
{
"family": "dollhousemcp",
"containerDefinitions": [{
"name": "dollhousemcp",
"image": "ghcr.io/mickdarling/dollhousemcp:latest",
"memory": 512,
"cpu": 256,
"environment": [
{"name": "NODE_ENV", "value": "production"},
{"name": "PERSONAS_DIR", "value": "/app/personas"},
{"name": "DOLLHOUSE_USER", "value": "production"}
]
}]
}
Google Cloud Run
gcloud run deploy dollhousemcp \
--image ghcr.io/mickdarling/dollhousemcp:latest \
--platform managed \
--region us-central1 \
--set-env-vars NODE_ENV=production,DOLLHOUSE_USER=production
Azure Container Instances
az container create \
--name dollhousemcp \
--resource-group myResourceGroup \
--image ghcr.io/mickdarling/dollhousemcp:latest \
--environment-variables NODE_ENV=production DOLLHOUSE_USER=production
🏗️ Project Structure
DollhouseMCP/
├── .github/
│ ├── actions/
│ │ └── validate-yaml/ # Reusable YAML validation action
│ └── workflows/
│ ├── cross-platform.yml # Cross-platform testing (Mac/Windows/Linux)
│ ├── claude.yml # Interactive Claude Code workflow
│ └── claude-code-review.yml # Automated PR review workflow
├── __tests__/
│ └── auto-update.test.ts # Comprehensive test suite (102 tests total)
├── src/
│ └── index.ts # Main MCP server (DollhouseMCPServer class)
├── dist/ # Compiled JavaScript (auto-generated)
├── personas/ # Local persona collection
│ ├── creative-writer.md # Enhanced with unique ID system
│ ├── technical-analyst.md
│ ├── eli5-explainer.md
│ ├── business-consultant.md
│ └── debug-detective.md
├── Dockerfile # Multi-stage Docker build
├── docker-compose.yml # Production and development configurations
├── .dockerignore # Docker build optimizations
├── package.json # Project config (dollhousemcp, AGPL-3.0)
├── tsconfig.json # TypeScript configuration
├── setup.sh # Automated installation script
├── LICENSE # AGPL-3.0 with platform stability terms
├── CONVERSATION_SUMMARY.md # Development session documentation
├── claude.md # Project context file
└── README.md # This file
📝 Creating Custom Personas
Enhanced Persona Format
Create .md files in the personas/ directory with this structure:
---
name: "Your Persona Name"
description: "Brief description of what this persona does"
unique_id: "auto-generated-if-missing"
author: "your-username"
triggers: ["keyword1", "keyword2"]
version: "1.0"
category: "creative"
age_rating: "all"
ai_generated: false
generation_method: "human"
price: "free"
license: "CC-BY-SA-4.0"
---
# Your Persona Name
Your persona instructions go here. This content defines how the AI should behave when this persona is activated.
## Response Style
- Communication guidelines
- Tone and approach
- Specific behaviors
## Key Techniques
- Problem-solving methods
- Interaction patterns
Metadata Fields
| Field | Required | Description |
|---|---|---|
name |
✅ | Display name for the persona |
description |
✅ | Brief description of purpose |
unique_id |
❌ | Auto-generated if missing |
author |
❌ | Creator username (uses environment or anonymous) |
category |
❌ | creative, professional, educational, gaming, personal |
triggers |
❌ | Keywords that suggest this persona |
version |
❌ | Version tracking |
age_rating |
❌ | all, 13+, 18+ |
ai_generated |
❌ | Boolean flag for AI-created content |
price |
❌ | "free" or monetary amount |
📚 Built-in Personas
| Persona | Purpose | Best For |
|---|---|---|
| Creative Writer | Imaginative storytelling and creative content | Brainstorming, creative writing, engaging narratives |
| Technical Analyst | Deep technical analysis and systematic problem-solving | Architecture decisions, debugging, technical docs |
| ELI5 Explainer | Simplifying complex topics for beginners | Teaching, onboarding, concept explanation |
| Business Consultant | Strategic business analysis and recommendations | Strategy planning, business decisions, market analysis |
| Debug Detective | Systematic debugging and troubleshooting | Bug hunting, system troubleshooting, root cause analysis |
🏪 Marketplace Integration
DollhouseMCP includes a complete GitHub-powered marketplace:
- Browse by Category: creative, professional, educational, gaming, personal
- Search Content: Find personas by keywords and descriptions
- One-Click Install: Download and integrate marketplace personas
- Community Submissions: Submit your personas via automated GitHub workflow
- Version Control: Full Git history for all marketplace content
💼 Business Model & Legal
Licensing
- Core Server: AGPL-3.0 (prevents proprietary competing platforms)
- Platform Terms: Creator-friendly 80/20 revenue split framework
- Persona Content: CC-BY-SA-4.0 for free personas
Platform Stability Commitments
- 90-day advance notice for monetization changes
- 12-month revenue sharing locks
- Full data portability rights
- Community advisory input on policy changes
🛠️ Development
Available Scripts
| Script | Description |
|---|---|
npm run build |
Compile TypeScript to JavaScript |
npm run start |
Run the compiled server |
npm run dev |
Run in development mode with auto-reload |
npm run clean |
Remove compiled files |
npm run rebuild |
Clean and rebuild the project |
npm run setup |
Install dependencies and build |
npm test |
Run the comprehensive test suite |
npm run test:coverage |
Run tests with coverage reporting |
Environment Variables
Customize server behavior with these environment variables:
# User Attribution
export DOLLHOUSE_USER="your-username" # User attribution for persona creation
export DOLLHOUSE_EMAIL="your-email" # Contact email (optional)
# Directory Configuration
export PERSONAS_DIR="/custom/path/to/personas" # Custom personas directory
# Auto-Update Configuration
export GITHUB_TOKEN="your-token" # For private repository access (optional)
# Development Configuration
export NODE_ENV="development" # Development mode
export DEBUG="dollhousemcp:*" # Debug logging (optional)
🔧 Troubleshooting
Common Issues
| Issue | Solution |
|---|---|
| Personas not loading | Check personas/ directory and file permissions |
| Server won't start | Run npm run rebuild to clean and rebuild |
| Marketplace not working | Check internet connection and GitHub API access |
| User identity not saving | Verify environment variables are set correctly |
Debug Steps
-
Check build status:
npm run build -
Verify persona files:
ls -la personas/*.md -
Test server startup:
node dist/index.js -
Validate configuration:
# Check Claude Desktop config cat ~/Library/Application\ Support/Claude/claude_desktop_config.json # Verify Node.js version (requires 18+) node --version # Check npm version npm --version -
Test auto-update system:
# Use within Claude Desktop check_for_updates # Check for available updates get_server_status # View system information -
Validate personas: Use the
reload_personastool to check for loading errors
🤝 Contributing
We welcome contributions! Here's how to help:
Adding Personas
- Fork the repository
- Create a new persona file in
personas/ - Follow the established format and naming conventions
- Test your persona thoroughly with
validate_personatool - Submit a pull request with a clear description
Community Contributions
- Create personas following the enhanced metadata format
- Test thoroughly with
validate_personatool - Submit via
submit_personatool for community review - Participate in GitHub discussions and issue reviews
Reporting Issues
Please include:
- Node.js version (
node --version) - Operating system and version
- Complete error messages
- Steps to reproduce the issue
- Relevant persona files (if applicable)
- Claude Desktop configuration (without sensitive paths)
Development Guidelines
- Follow TypeScript best practices
- Maintain existing code style and patterns
- Add comprehensive error handling
- Update documentation for new features
- Test thoroughly across platforms before submitting PRs
- Include tests for new functionality
- Follow semantic versioning for releases
Development Workflow
# Fork and clone
git clone https://github.com/your-username/DollhouseMCP.git
cd DollhouseMCP
# Install dependencies
npm install
# Create feature branch
git checkout -b feature/your-feature-name
# Make changes and test
npm run build
npm test
# Commit and push
git commit -m "feat: your feature description"
git push origin feature/your-feature-name
# Submit pull request
📄 API Reference
MCP Tool Specifications
Each tool follows the MCP specification:
interface DollhouseTool {
name: string;
description: string;
inputSchema: {
type: "object";
properties: Record<string, any>;
required?: string[];
};
}
Tool Categories
Core Persona Management
// list_personas - No parameters
// activate_persona - { persona: string }
// get_active_persona - No parameters
// deactivate_persona - No parameters
// get_persona_details - { persona: string }
// reload_personas - No parameters
Marketplace Integration
// browse_marketplace - { category?: string }
// search_marketplace - { query: string }
// get_marketplace_persona - { path: string }
// install_persona - { path: string }
// submit_persona - { persona: string }
User Identity Management
// set_user_identity - { username: string }
// get_user_identity - No parameters
// clear_user_identity - No parameters
Chat-Based Management
// create_persona - { name: string, description: string, category?: string, instructions: string }
// edit_persona - { persona: string, field: string, value: string }
// validate_persona - { persona: string }
Auto-Update System
// check_for_updates - No parameters
// update_server - { confirm: boolean }
// rollback_update - { confirm: boolean }
// get_server_status - No parameters
Error Handling
The server provides detailed error messages for:
- Invalid persona identifiers - Clear suggestions for valid names
- File system issues - Permission and path resolution errors
- Malformed persona files - YAML parsing and validation errors
- Network errors - GitHub API and marketplace connectivity issues
- Runtime errors - Server startup and operation failures
Response Formats
All responses follow consistent patterns:
- Success responses: Include requested data and operation status
- Error responses: Include error type, message, and suggested resolution
- Progress indicators: Step-by-step feedback for long operations
- Validation results: Detailed reports with recommendations
📄 License
This project is licensed under the AGPL-3.0 License with Platform Stability Commitments - see the LICENSE file for details.
Platform Stability Guarantees:
- 90-day advance notice for policy changes
- 12-month revenue sharing locks
- Full data portability rights
- Community advisory input
🏷️ Version History
Unreleased (Current)
- ✅ Persona active indicator system with 2 new MCP tools (Issue #31)
- ✅ 102 comprehensive tests covering all functionality
- ✅ 23 total MCP tools including configure_indicator and get_indicator_config
- ✅ Enhanced documentation with Claude Desktop setup clarifications
v1.1.0 - July 4, 2025
- ✅ Platform-specific badges for Windows, macOS, Linux visibility
- ✅ GitHub Project management with issue templates and milestones
- ✅ ARM64 Docker fix switching from Alpine to Debian base images
- ✅ 100% workflow reliability (except Docker ARM64)
- ✅ First GitHub release with CHANGELOG.md
- ✅ 21 total MCP tools at time of release
Phase 2B+ - July 3, 2025
- ✅ Enterprise-grade auto-update system with 4 new MCP tools
- ✅ 50 comprehensive tests covering all functionality
- ✅ Security hardening - eliminated all command injection vulnerabilities
- ✅ Cross-platform support - Windows, macOS, Linux with CI/CD testing
- ✅ Docker containerization with production and development configurations
- ✅ 21 total MCP tools with backup/rollback and dependency validation
Phase 2B - July 1-2, 2025
- ✅ Complete chat-based persona management
- ✅ GitHub marketplace integration
- ✅ User identity and attribution system
- ✅ Real-time validation and editing
- ✅ Enterprise-grade GitHub Actions security
Phase 1 - July 1, 2025
- ✅ Fresh AGPL-3.0 licensed repository
- ✅ Enhanced unique ID system
- ✅ Anonymous user support
- ✅ Marketplace-ready metadata schema
🎭 Transform your AI interactions with the power of personas
For support, please open an issue or visit our marketplace.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.