
MCP Launch Assistant
Manages the lifecycle of MCP services with health monitoring, auto-restart capabilities, and process tracking. Enables users to start, stop, restart, and monitor MCP services through natural language commands.
README
MCP Launch Assistant
⚠️ DEPRECATED: This service has been replaced by
mcp-services-assistant
which provides better separation of concerns and improved LaunchAgent management. Please migrate to the new service.
Service lifecycle management for the MCP ecosystem.
Overview
The Launch Assistant manages the lifecycle of all MCP services in the ecosystem with integrated health monitoring and auto-restart capabilities.
Features
- Service Management: Start, stop, restart, and monitor services
- Health Monitoring: Automatic health checks every 30 seconds for services with ports
- Auto-Restart: Automatically restart failed services (up to 3 attempts by default)
- Process Tracking: Track PIDs, uptime, restart counts, and exit codes
- Log Capture: Capture stdout/stderr and keep last 100 lines in memory
- Graceful Shutdown: 10-second grace period before force-killing
- Service Discovery: Auto-discover services from registry
- LaunchAgent Support: macOS system integration for auto-start
Installation
npm install
Usage
As MCP Server (Claude Desktop)
Configure in Claude Desktop settings:
{
"mcpServers": {
"launch": {
"command": "node",
"args": ["/path/to/mcp-launch-assistant/index.js"]
}
}
}
As HTTP Service
npm start
The service will be available at http://localhost:9102
Available Tools
list_services
List all registered services and their status.
await client.callTool({
name: 'list_services',
arguments: {
category: 'platform' // optional: filter by category
}
});
start_service
Start a specific service.
await client.callTool({
name: 'start_service',
arguments: {
service: 'mcp-gateway-assistant',
mode: 'dual' // 'mcp', 'http', or 'dual'
}
});
stop_service
Stop a running service.
await client.callTool({
name: 'stop_service',
arguments: {
service: 'mcp-gateway-assistant'
}
});
restart_service
Restart a service.
await client.callTool({
name: 'restart_service',
arguments: {
service: 'mcp-gateway-assistant'
}
});
service_status
Get detailed status of a service including health and logs.
await client.callTool({
name: 'service_status',
arguments: {
service: 'mcp-gateway-assistant'
}
});
// Returns:
{
service: 'mcp-gateway-assistant',
status: 'running',
pid: 12345,
startedAt: '2025-01-13T10:00:00Z',
uptime: 3600000,
restarts: 0,
health: {
status: 'healthy',
responseTime: '45ms'
},
recentLogs: [...]
}
setup_launchagent
Create a LaunchAgent for automatic startup (macOS only).
await client.callTool({
name: 'setup_launchagent',
arguments: {
service: 'mcp-gateway-assistant'
}
});
Configuration
Constructor Options
const launcher = new McpLaunchAssistant({
healthCheckInterval: 30000, // Check health every 30 seconds
autoRestart: true, // Auto-restart failed services
maxRestarts: 3, // Max restart attempts
restartDelay: 5000, // Wait 5 seconds before restart
gracefulShutdownTimeout: 10000 // 10 seconds for graceful shutdown
});
Environment Variables
SERVICE_PORT=9103 # Launch assistant port
SERVICE_MODE=dual # Default service mode
LOG_LEVEL=info # Logging level
NODE_ENV=production # Environment
Data Storage
- Service metadata:
~/Documents/mcp-assistant/data/launch-assistant/services.json
- Process info:
~/Documents/mcp-assistant/data/launch-assistant/processes.json
- Logs:
~/Documents/mcp-assistant/logs/launch-assistant.log
Development
# Run tests
npm test
# Development mode
npm run dev
Events
The launcher emits these events:
serviceStarted
- Service successfully startedserviceStopped
- Service stoppedserviceRestarted
- Service restarted after failureserviceUnhealthy
- Health check failedserviceExited
- Process exitedserviceLog
- New log entryserviceError
- Error log entry
Troubleshooting
Service won't start
- Check if port is already in use
- Verify service directory exists
- Check NODE_PATH environment variable
- Review service logs with service_status
Health checks failing
- Ensure service implements /health endpoint
- Check if service is actually listening on configured port
- Increase health check timeout if service is slow
Auto-restart not working
- Check if max restarts exceeded
- Verify autoRestart option is enabled
- Check logs for crash reasons
LaunchAgent not working
- Ensure proper permissions
- Check ~/Library/LaunchAgents/
- Review system logs with
log show --predicate 'subsystem == "com.mcpassistant"'
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.