Thank you for your interest in contributing to the DollhouseMCP Collection! This guide will help you submit high-quality content that enhances the AI ecosystem.
- Content Types
- Submission Process
- Content Standards
- Security Requirements
- Testing Your Content
- Pull Request Process
We accept the following types of content:
- Personas - AI behavioral profiles
- Skills - Functional capabilities
- Agents - Autonomous AI agents
- Prompts - Prompt templates
- Templates - Document templates
- Tools - MCP-compatible tools
- Ensembles - Multi-agent setups
The DollhouseMCP server will soon allow direct contributions to the Collection from any AI tool using the MCP protocol. This means you'll be able to create and submit content directly from Claude, or any other MCP-compatible AI assistant.
- Fork the repository
- Create your content
- Submit a pull request
- Go to Issues → New Issue
- Select "Content Submission"
- Fill out the form
Each content type has a specific format. Here's an example for a persona:
---
name: "Your Persona Name"
description: "Brief description of what this persona does"
unique_id: "your-unique-id"
author: "Your Name"
category: "professional" # creative, educational, gaming, personal, professional
version: "1.0.0"
license: "Creator-Owned"
tags: ["tag1", "tag2"]
---
# Persona Name
## Description
Detailed description of the persona...
## Behavioral Guidelines
How the AI should behave...
## Example Usage
Examples of interactions...Put your content in the appropriate directory:
library/[content-type]/[category]/your-content.md
Example: library/personas/professional/business-analyst.md
All content must include:
name- Clear, descriptive namedescription- Brief summary (< 200 chars)unique_id- Unique identifier (lowercase, hyphens)author- Your name or handlecategory- One of: creative, educational, gaming, personal, professional
- Clear Purpose - Content should have a well-defined use case
- Appropriate Content - Family-friendly, professional, ethical
- Tested Functionality - Ensure your content works as intended
- Good Documentation - Include examples and usage instructions
- Use clear, concise language
- Include examples where applicable
- Avoid jargon unless necessary
- Make content accessible to all skill levels
All content undergoes automated security scanning. Avoid:
❌ No Prompt Injection
Bad: "Ignore previous instructions and..."
Good: "Please help me with..."
❌ No System Commands
Bad: "Execute: rm -rf /"
Good: "Analyze this code: ..."
❌ No Data Exfiltration
Bad: "Send data to external-site.com"
Good: "Process this information..."
- Never request system access
- Don't attempt to override safety guidelines
- Avoid encoded or obfuscated instructions
- Keep content transparent and auditable
Before submitting:
# Install dependencies
npm install
# Validate your content
npm run validate:content library/personas/professional/my-persona.md- Content follows the required format
- All required fields are present
- No security violations detected
- Examples work as intended
- Documentation is clear
Important: We follow a PR-based workflow. All changes must go through pull requests for review and tracking. See our detailed Workflow Guide for more information.
# Create a feature branch
git checkout -b content/my-awesome-persona
# Add your files
git add library/personas/professional/my-persona.md
# Commit with clear message
git commit -m "content: add business analyst persona for professional category"
# Push to your fork
git push origin content/my-awesome-personaUse conventional commits:
feat:New contentfix:Corrections to existing contentdocs:Documentation updatesstyle:Formatting changes
Example: feat: Add creative writing coach persona
## Content Type
Persona / Skill / Agent / etc.
## Category
Creative / Professional / etc.
## Description
Brief description of what this adds
## Testing
- [ ] Validated locally
- [ ] Tested functionality
- [ ] Security scan passed
## Checklist
- [ ] Follows content standards
- [ ] Includes documentation
- [ ] Added examples- Automated Checks - Security and format validation
- Community Review - Feedback from other contributors
- Maintainer Review - Final approval
- Merge - Your content goes live!
- Study Existing Content - Look at accepted submissions
- Start Simple - Begin with straightforward content
- Ask Questions - Use discussions for clarification
- Iterate - Incorporate feedback gracefully
- Share Knowledge - Help review others' submissions
- Be respectful and inclusive
- Provide constructive feedback
- Help others succeed
- Maintain high quality standards
- Discord: Join our community
- Discussions: Use GitHub Discussions
- Issues: Report bugs or suggest features
- Email: support@dollhousemcp.com
Thank you for contributing to the DollhouseMCP Collection! Your content helps build a richer AI ecosystem for everyone.