Environments

Environments separate development, staging, and production deployments.

Environment Types

Development

For local development and testing:

swiftclaw deploy my-agent --env development

Characteristics:

Debug logging enabled
Lower rate limits
Test API keys
Faster deployment

Staging

For pre-production testing:

swiftclaw deploy my-agent --env staging

Characteristics:

Production-like configuration
Isolated from production data
Full testing capabilities
Separate API keys

Production

For live deployments:

swiftclaw deploy my-agent --env production

Characteristics:

Optimized performance
High availability
Production API keys
Monitoring and alerts

Environment Configuration

Environment Variables

# Development
export SWIFTCLAW_ENV=development
export SWIFTCLAW_API_KEY=sk_test_abc123...

# Production
export SWIFTCLAW_ENV=production
export SWIFTCLAW_API_KEY=sk_live_xyz789...

Configuration Files

// config/development.json
{
  "model": "gpt-3.5-turbo",
  "debug": true,
  "log_level": "debug"
}

// config/production.json
{
  "model": "gpt-4",
  "debug": false,
  "log_level": "info"
}

Environment Isolation

Each environment has:

Separate Instances: Independent agent deployments
Isolated Data: No cross-environment data access
Different API Keys: Environment-specific authentication
Independent Scaling: Per-environment resource allocation

Promoting Between Environments

Manual Promotion

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

# Test in staging
swiftclaw test my-agent --env staging

# Promote to production
swiftclaw deploy my-agent --env production --version staging-v1.2.0

Automated Promotion

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

jobs:
  deploy-staging:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v3
      - name: Deploy to Staging
        run: swiftclaw deploy my-agent --env staging

  deploy-production:
    needs: deploy-staging
    runs-on: ubuntu-latest
    steps:
      - name: Deploy to Production
        run: swiftclaw deploy my-agent --env production

Environment-Specific Settings

Model Configuration

# Development: Use faster, cheaper model
swiftclaw config set model gpt-3.5-turbo --env development

# Production: Use powerful model
swiftclaw config set model gpt-4 --env production

Scaling Configuration

# Development: Minimal resources
swiftclaw config set min-instances 1 --env development

# Production: High availability
swiftclaw config set min-instances 3 --env production

Memory Configuration

# Development: Short TTL
swiftclaw config set memory-ttl 1h --env development

# Production: Long TTL
swiftclaw config set memory-ttl 30d --env production

Environment Variables

Required Variables

SWIFTCLAW_ENV=production
SWIFTCLAW_API_KEY=sk_live_abc123...
SWIFTCLAW_WORKSPACE=my-workspace

Optional Variables

SWIFTCLAW_BASE_URL=https://api.swiftclaw.io
SWIFTCLAW_TIMEOUT=30
SWIFTCLAW_MAX_RETRIES=3

Best Practices

1. Use Separate API Keys

# Development
SWIFTCLAW_API_KEY=sk_test_dev123...

# Production
SWIFTCLAW_API_KEY=sk_live_prod456...

2. Test in Staging First

# Always test in staging
swiftclaw deploy my-agent --env staging
swiftclaw test my-agent --env staging

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

3. Monitor Each Environment

# Set up environment-specific alerts
swiftclaw alerts create \
  --agent my-agent \
  --env production \
  --metric error-rate \
  --threshold 1

4. Use Environment-Specific Configurations

# Load environment-specific config
swiftclaw deploy my-agent \
  --env production \
  --config config/production.json

Never use production keys in development: Always use test keys for development and staging.

Environment Status

Check environment status:

# View all environments
swiftclaw envs list my-agent

# View specific environment
swiftclaw envs status my-agent --env production

Next Steps

Environments

Environments separate development, staging, and production deployments.

Environment Types

Development

For local development and testing:

swiftclaw deploy my-agent --env development

Characteristics:

Debug logging enabled
Lower rate limits
Test API keys
Faster deployment

Staging

For pre-production testing:

swiftclaw deploy my-agent --env staging

Characteristics:

Production-like configuration
Isolated from production data
Full testing capabilities
Separate API keys

Production

For live deployments:

swiftclaw deploy my-agent --env production

Characteristics:

Optimized performance
High availability
Production API keys
Monitoring and alerts

Environment Configuration

Environment Variables

# Development
export SWIFTCLAW_ENV=development
export SWIFTCLAW_API_KEY=sk_test_abc123...

# Production
export SWIFTCLAW_ENV=production
export SWIFTCLAW_API_KEY=sk_live_xyz789...

Configuration Files

// config/development.json
{
  "model": "gpt-3.5-turbo",
  "debug": true,
  "log_level": "debug"
}

// config/production.json
{
  "model": "gpt-4",
  "debug": false,
  "log_level": "info"
}

Environment Isolation

Each environment has:

Separate Instances: Independent agent deployments
Isolated Data: No cross-environment data access
Different API Keys: Environment-specific authentication
Independent Scaling: Per-environment resource allocation

Promoting Between Environments

Manual Promotion

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

# Test in staging
swiftclaw test my-agent --env staging

# Promote to production
swiftclaw deploy my-agent --env production --version staging-v1.2.0

Automated Promotion

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

jobs:
  deploy-staging:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v3
      - name: Deploy to Staging
        run: swiftclaw deploy my-agent --env staging

  deploy-production:
    needs: deploy-staging
    runs-on: ubuntu-latest
    steps:
      - name: Deploy to Production
        run: swiftclaw deploy my-agent --env production

Environment-Specific Settings

Model Configuration

# Development: Use faster, cheaper model
swiftclaw config set model gpt-3.5-turbo --env development

# Production: Use powerful model
swiftclaw config set model gpt-4 --env production

Scaling Configuration

# Development: Minimal resources
swiftclaw config set min-instances 1 --env development

# Production: High availability
swiftclaw config set min-instances 3 --env production

Memory Configuration

# Development: Short TTL
swiftclaw config set memory-ttl 1h --env development

# Production: Long TTL
swiftclaw config set memory-ttl 30d --env production

Environment Variables

Required Variables

SWIFTCLAW_ENV=production
SWIFTCLAW_API_KEY=sk_live_abc123...
SWIFTCLAW_WORKSPACE=my-workspace

Optional Variables

SWIFTCLAW_BASE_URL=https://api.swiftclaw.io
SWIFTCLAW_TIMEOUT=30
SWIFTCLAW_MAX_RETRIES=3

Best Practices

1. Use Separate API Keys

# Development
SWIFTCLAW_API_KEY=sk_test_dev123...

# Production
SWIFTCLAW_API_KEY=sk_live_prod456...

2. Test in Staging First

# Always test in staging
swiftclaw deploy my-agent --env staging
swiftclaw test my-agent --env staging

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

3. Monitor Each Environment

# Set up environment-specific alerts
swiftclaw alerts create \
  --agent my-agent \
  --env production \
  --metric error-rate \
  --threshold 1

4. Use Environment-Specific Configurations

# Load environment-specific config
swiftclaw deploy my-agent \
  --env production \
  --config config/production.json

Never use production keys in development: Always use test keys for development and staging.

Environment Status

Check environment status:

# View all environments
swiftclaw envs list my-agent

# View specific environment
swiftclaw envs status my-agent --env production

Environments

On this page

Environments

On this page