xRegistry for AI

Automated registry for AI protocol metadata with issue-driven submissions and static site deployment.

What This Repository Does

🎫 Issue-Driven Registration

Submit AI protocol implementations through GitHub issues:

• Agent-to-Agent (A2A) registrations via A2A issue template
• Model Context Protocol (MCP) registrations via MCP issue template
• Automated validation and ingestion from remote repositories

🔄 Automated Workflows

• Issue Processing: Validates submissions and fetches metadata from remote repos
• Schema Generation: Auto-generates unified OpenAPI, JSON Schema, and Avro schemas
• Static Site Build: Packages registry data into deployable static site
• Azure Deployment: Deploys to Azure Static Web Apps with Angular SPA

📡 Deployed Services

Live Registry Site

• Primary: xregistry-ai.azurestaticapps.net (Azure Static Web Apps)
• GitHub Pages: clemensv.github.io/xregistry-ai

API Endpoints

• /registry/ - Browse all registered protocols
• /registry/mcpproviders/ - MCP server implementations
• /registry/agentcardproviders/ - A2A agent implementations
• /openapi/openapi.json - Unified OpenAPI specification
• /schema/json-schema.json - Unified JSON Schema
• /schema/avro-schema.json - Unified Avro Schema

Search Indices

• /flex.json files for protocol-specific search
• Full-text search across all registered implementations

Quick Start

Submit Your Protocol Implementation

1. Ensure your repo has the required metadata file (mcp.json or agent.json)
2. Open an issue using the appropriate template
3. Automated validation will check your submission
4. Upon approval, your implementation is added to the registry

Local Development

git clone <repository-url>
cd xregistry-ai
pip install -r requirements.txt
cd xreg && ./build-all.sh

Supported Protocols

• MCP (Model Context Protocol): Server implementations with mcp.json
• A2A (Agent-to-Agent): Agent cards with agent.json
• Extensible: Add new protocols via model definitions

Architecture

• Registry Backend: Custom xRegistry server (containerized)
• Frontend: Angular SPA (xregistry-viewer)
• Deployment: Azure Static Web Apps + GitHub Actions
• Search: Flex search indices with full-text capabilities