MoM (Mule of Mules)

MoM is a Ruby on Rails manager for Mule AI agents. It provides centralized management and coordination of multiple Mule instances through both REST API and gRPC interfaces.

Features

• Client Management: Register, monitor, and manage multiple Mule AI clients
• Workflow Orchestration: Execute workflows across different Mule instances
• gRPC Integration: High-performance gRPC server for real-time communication
• REST API: HTTP endpoints for web-based integrations
• Status Monitoring: Track client health and workflow execution status

Quick Start

1. Install dependencies:

   bundle install

2. Generate gRPC code:

   rake generate_grpc

3. Setup database:

   rails db:create db:migrate

4. Start the server:

   rails server

The gRPC server will start automatically on port 50051 (configurable via GRPC_PORT environment variable).

API Endpoints

REST API

• GET /api/v1/mule_clients - List all registered clients
• POST /api/v1/mule_clients - Register a new client
• GET /api/v1/mule_clients/:id - Get client details
• DELETE /api/v1/mule_clients/:id - Deregister a client
• POST /api/v1/mule_clients/:id/execute_workflow - Execute workflow on client
• GET /api/v1/mule_clients/:id/status - Get client status and running workflows

gRPC Service

The MuleManagerService provides equivalent functionality through gRPC:

• RegisterMuleClient - Register a new Mule client
• GetMuleClients - List all registered clients
• ExecuteWorkflowOnClient - Execute a workflow on a specific client
• GetClientStatus - Get client status and running workflows
• DeregisterMuleClient - Remove a client from management

Configuration

Environment variables:

• GRPC_PORT - gRPC server port (default: 50051)
• GRPC_HOST - gRPC server host (default: 0.0.0.0)
• DB_USERNAME - Database username
• DB_PASSWORD - Database password
• DB_HOST - Database host

Development

Generate gRPC code after modifying the protobuf definition:

rake generate_grpc

Architecture

MoM acts as a central coordinator for multiple Mule instances:

1. Client Registration: Mule instances register themselves with MoM
2. Workflow Distribution: MoM routes workflow requests to appropriate clients
3. Status Monitoring: Track execution status and client health
4. Load Balancing: Distribute work across available clients (future feature)

This enables scalable AI agent orchestration across multiple machines or containers.