Deployment

ContextVM Deployment

Deploy ContextVM servers and clients in production environments.

Environment Variables

Required

Optional

Docker Deployment

Basic Server Container

FROM oven/bun:alpine

WORKDIR /app

COPY package.json bun.lock ./
RUN bun install --frozen-lockfile

COPY . .

ENV SERVER_PRIVATE_KEY=""
ENV RELAYS="wss://relay.contextvm.org"
ENV LOG_LEVEL="info"

EXPOSE 3000

CMD ["bun", "run", "server.ts"]

Docker Compose

version: '3.8'
services:
  cvm-server:
    build: .
    environment:
      - SERVER_PRIVATE_KEY=${SERVER_PRIVATE_KEY}
      - RELAYS=wss://relay.contextvm.org
      - LOG_LEVEL=info
      - ENCRYPTION_MODE=optional
    restart: unless-stopped
    logging:
      driver: 'json-file'
      options:
        max-size: '10m'
        max-file: '3'

See assets/docker-compose.yml for complete example.

Relay Configuration

Recommended Public Relays

wss://relay.contextvm.org
wss://nos.lol

Production Considerations

• Use 3-5 relays for redundancy
• Include at least 2 geographically distributed
• Monitor relay uptime
• Have backup relay list ready

Private Relay Setup

For sensitive deployments:

const relayPool = new ApplesauceRelayPool(['wss://private-relay.your-domain.com']);

Security Best Practices

Key Management

✓ DO:

• Store keys in environment variables
• Use secret management, if the security of the server requires it (AWS Secrets Manager, HashiCorp Vault)

✗ DON'T:

• Hardcode keys in source
• Commit keys to version control
• Log private keys
• Share keys between services

Access Control

// Whitelist specific clients
const transport = new NostrServerTransport({
  signer,
  relayHandler: relayPool,
  allowedPublicKeys: [process.env.CLIENT_1_PUBKEY!, process.env.CLIENT_2_PUBKEY!],
});

Health Checks

Server Health Check

async function healthCheck(server: McpServer): Promise<boolean> {
  try {
    // Check if server responds to ping
    await server.ping();
    return true;
  } catch {
    return false;
  }
}

Docker Health Check

HEALTHCHECK --interval=30s --timeout=10s --start-period=5s --retries=3 \
  CMD bun run healthcheck.ts || exit 1

Monitoring

Structured Logging

import { createLogger } from '@contextvm/sdk/core';

const logger = createLogger('server');

// Production log format
logger.info('request.completed', {
  module: 'server',
  method: 'tools/call',
  tool: 'echo',
  clientPubkey: pubkey.slice(0, 8) + '...',
  durationMs: 45,
});

Metrics to Track

• Request rate and latency
• Error rate by type
• Active connections
• Relay connection status
• Event publish/subscribe rates

See references/monitoring.md for detailed monitoring setup.

Production Checklist

• Keys in environment/secrets manager
• Logging configured appropriately
• Health checks implemented
• Error handling in place
• Graceful shutdown handling
• Resource limits set (Docker)
• Monitoring/alerting configured