MCP Server For Creating Registry Entries

[Brandon-Anubis]

Created an MCP Server for Quickly Converting any MCP to ToolHive Registry According To Schema.

Overview

The MCP Registry Converter, a fully-functional MCP (Model Context Protocol) server that automates the conversion of MCP server repositories into ToolHive-compatible registry entries. This tool significantly reduces the manual effort required to add new servers to the registry, transforming a 15-20 minute manual process into a sub-minute automated workflow.

What This Adds

Core MCP Server Implementation

The converter provides two powerful tools and one prompt accessible via MCP protocol:

1. convert_repo_to_toolhive

• Accepts a GitHub repository URL as input
• Fetches and analyzes the README.md file
• Extracts server metadata using GitHub API (stars, forks, last updated)
• Intelligently parses tool lists and descriptions
• Optionally accepts AI-suggested tags and descriptions from the enhance_registry_entry prompt
• Generates properly formatted ToolHive spec.yaml output
• Validates the result against the ToolHive schema

2. validate_toolhive_yaml

• Validates YAML entries against the ToolHive registry schema
• Checks required fields (name, description, transport, repository_url)
• Verifies field values (transport types, tier, status)
• Reports detailed validation errors

3. enhance_registry_entry (MCP Prompt)

• Analyzes an MCP server repository to suggest tags and refined descriptions
• Uses the AI client running the server (Claude Desktop, local AI, etc.)
• No external API calls required

Key Features

• GitHub API Integration: Fetches README content and repository metadata (stars, forks, topics, last updated timestamp)
• AI Enhancement via MCP Prompts: Uses the enhance_registry_entry prompt to let any MCP-compatible AI client (Claude Desktop, local AI via LMStudio/Ollama, etc.) provide suggestions directly - no external API keys needed
• Local AI Support: Works seamlessly with local AI models through the MCP protocol
• Smart Extraction: Parses README structure to identify tools, transport type, and configuration
• YAML Generation: Produces properly formatted spec.yaml files ready for the registry
• Schema Validation: Ensures all outputs conform to ToolHive requirements
• Graceful Degradation: Works without AI or even without GitHub API access (generates minimal valid entries)
• Comprehensive Error Handling: Provides clear error messages and validation feedback
• Privacy-Friendly: All AI analysis happens through your chosen client - no external API calls

Usage Examples

With Claude Desktop (AI-enhanced):

"Use the enhance_registry_entry prompt for https://github.com/mongodb-js/mongodb-mcp-server, then convert it to a ToolHive registry entry"

With Local AI (LMStudio/Ollama):
Works the same way - your local AI provides analysis through the MCP prompt!

Basic conversion (no AI):

"Convert the GitHub repository at https://github.com/mongodb-js/mongodb-mcp-server to a ToolHive registry entry"

With ToolHive:

thv install mcp-registry-converter
thv mcp run mcp-registry-converter

Direct Usage:

cd mcp-registry-converter
npm install && npm run build && npm start

Documentation

Comprehensive documentation included:

• README.md - Complete technical documentation
• EXAMPLES.md - Real-world usage examples and patterns
• INTEGRATION.md - Deployment and integration guide
• MCP_CONVERTER_GUIDE.md - End-user guide
• IMPLEMENTATION_SUMMARY.md - Technical implementation overview
• QUICK_REFERENCE.md - Quick reference card

Deployment & CI/CD

• Docker Support: Production-ready Dockerfile with security hardening (non-root user)
• GitHub Actions: Automated build, test, and publish workflow
• Container Registry: Automatic publishing to GitHub Container Registry on merge
• Registry Entry: The converter itself is registered in the ToolHive registry

Testing

All integration tests passing (5/5):

• ✅ README extraction and parsing
• ✅ YAML validation (valid and invalid cases)
• ✅ Spec generation and formatting
• ✅ Transport type validation
• ✅ Build and compilation

Impact

This tool will:

• Accelerate Onboarding: New MCP servers can be added to the registry in seconds instead of minutes
• Improve Quality: AI-enhanced suggestions (via MCP prompts) and automatic validation ensure consistent, high-quality entries
• Reduce Errors: Automated extraction and validation minimize human error
• Enable Bulk Import: Multiple repositories can be converted efficiently
• Lower Barriers: Makes it easier for community members to contribute new servers
• Support Local AI: Works with any MCP-compatible AI client including local models (LMStudio, Ollama, etc.)
• Privacy-Friendly: No external API calls for AI features - all analysis through your chosen AI client

Files Changed

• mcp-registry-converter/ - Complete TypeScript MCP server implementation
• registry/mcp-registry-converter/ - Registry entry for the converter itself
• .github/workflows/build-mcp-converter.yml - CI/CD automation
• Documentation files (README, EXAMPLES, guides)

Environment Variables

Optional but recommended:

• GITHUB_TOKEN - Increases GitHub API rate limit from 60/hour to 5000/hour