Production Deployment

Learn how to deploy SwiftClaw agents to production with confidence.

Pre-Deployment Checklist

Before deploying to production:

Test agent thoroughly in development
Configure environment variables
Set up monitoring and alerts
Review security settings
Configure scaling policies
Set up backup and recovery

Deployment Methods

CLI Deployment

Deploy using the SwiftClaw CLI:

# Deploy to production
swiftclaw deploy my-agent --env production

# Deploy with specific configuration
swiftclaw deploy my-agent \
  --env production \
  --config production.json \
  --region us-east-1

Dashboard Deployment

Navigate to your agent in the dashboard
Click "Deploy to Production"
Review configuration
Confirm deployment

CI/CD Deployment

Integrate with your CI/CD pipeline:

# .github/workflows/deploy.yml
name: Deploy Agent
on:
  push:
    branches: [main]

jobs:
  deploy:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v3
      - name: Deploy to SwiftClaw
        run: |
          swiftclaw login --token ${{ secrets.SWIFTCLAW_TOKEN }}
          swiftclaw deploy my-agent --env production

Environment Configuration

Production Settings

{
  "environment": "production",
  "model": {
    "primary": "gpt-4",
    "fallbacks": ["claude-3-sonnet"]
  },
  "scaling": {
    "minInstances": 2,
    "maxInstances": 10,
    "targetCPU": 70
  },
  "monitoring": {
    "enabled": true,
    "logLevel": "info",
    "metrics": true
  }
}

Health Checks

SwiftClaw automatically performs health checks:

Startup Check: Verifies agent starts successfully
Liveness Check: Ensures agent is running
Readiness Check: Confirms agent can handle requests

Configure custom health checks:

@agent.health_check
async def custom_health():
    # Check database connection
    if not await db.ping():
        return {"status": "unhealthy", "reason": "database"}
    
    # Check external API
    if not await api.ping():
        return {"status": "unhealthy", "reason": "api"}
    
    return {"status": "healthy"}

Zero-Downtime Deployment

SwiftClaw uses rolling deployments:

New version deployed alongside old version
Health checks verify new version
Traffic gradually shifted to new version
Old version terminated after full migration

Automatic Rollback: If health checks fail, deployment automatically rolls back to previous version.

Post-Deployment Verification

Verify deployment success:

# Check deployment status
swiftclaw status my-agent

# View recent logs
swiftclaw logs my-agent --tail 100

# Test agent endpoint
curl https://api.swiftclaw.io/agents/my-agent/invoke \
  -H "Authorization: Bearer $TOKEN" \
  -d '{"message": "test"}'

Next Steps

Production Deployment

Learn how to deploy SwiftClaw agents to production with confidence.

Pre-Deployment Checklist

Before deploying to production:

Test agent thoroughly in development
Configure environment variables
Set up monitoring and alerts
Review security settings
Configure scaling policies
Set up backup and recovery

Deployment Methods

CLI Deployment

Deploy using the SwiftClaw CLI:

# Deploy to production
swiftclaw deploy my-agent --env production

# Deploy with specific configuration
swiftclaw deploy my-agent \
  --env production \
  --config production.json \
  --region us-east-1

Dashboard Deployment

Navigate to your agent in the dashboard
Click "Deploy to Production"
Review configuration
Confirm deployment

CI/CD Deployment

Integrate with your CI/CD pipeline:

# .github/workflows/deploy.yml
name: Deploy Agent
on:
  push:
    branches: [main]

jobs:
  deploy:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v3
      - name: Deploy to SwiftClaw
        run: |
          swiftclaw login --token ${{ secrets.SWIFTCLAW_TOKEN }}
          swiftclaw deploy my-agent --env production

Environment Configuration

Production Settings

{
  "environment": "production",
  "model": {
    "primary": "gpt-4",
    "fallbacks": ["claude-3-sonnet"]
  },
  "scaling": {
    "minInstances": 2,
    "maxInstances": 10,
    "targetCPU": 70
  },
  "monitoring": {
    "enabled": true,
    "logLevel": "info",
    "metrics": true
  }
}

Health Checks

SwiftClaw automatically performs health checks:

Startup Check: Verifies agent starts successfully
Liveness Check: Ensures agent is running
Readiness Check: Confirms agent can handle requests

Configure custom health checks:

@agent.health_check
async def custom_health():
    # Check database connection
    if not await db.ping():
        return {"status": "unhealthy", "reason": "database"}
    
    # Check external API
    if not await api.ping():
        return {"status": "unhealthy", "reason": "api"}
    
    return {"status": "healthy"}

Zero-Downtime Deployment

SwiftClaw uses rolling deployments:

New version deployed alongside old version
Health checks verify new version
Traffic gradually shifted to new version
Old version terminated after full migration

Automatic Rollback: If health checks fail, deployment automatically rolls back to previous version.

Post-Deployment Verification

Verify deployment success:

# Check deployment status
swiftclaw status my-agent

# View recent logs
swiftclaw logs my-agent --tail 100

# Test agent endpoint
curl https://api.swiftclaw.io/agents/my-agent/invoke \
  -H "Authorization: Bearer $TOKEN" \
  -d '{"message": "test"}'

Production Deployment

Production Deployment

Pre-Deployment Checklist

Deployment Methods

CLI Deployment

Dashboard Deployment

CI/CD Deployment

Environment Configuration

Production Settings

Health Checks

Zero-Downtime Deployment

Post-Deployment Verification

Next Steps

On this page

Production Deployment

Production Deployment

Pre-Deployment Checklist

Deployment Methods

CLI Deployment

Dashboard Deployment

CI/CD Deployment

Environment Configuration

Production Settings

Health Checks

Zero-Downtime Deployment

Post-Deployment Verification

Next Steps

On this page