codedeploy

name	codedeploy
description	Manage AWS CodeDeploy applications, deployment groups, deployments, and rollbacks via AWS CLI.
metadata	{"openclaw":{"emoji":"🚀","requires":{"bins":["aws"]}}}

AWS CodeDeploy

Use this skill for deployment automation: managing applications and deployment groups, triggering and monitoring deployments, configuring rollback and deployment strategies, and managing on-premises instances.

Prerequisites

• AWS CLI v2 configured with valid credentials
• IAM permissions: codedeploy:* for full access, or scoped policies
• CodeDeploy agent installed on target EC2 instances
• Service role with appropriate permissions for the compute platform (EC2, Lambda, or ECS)

Common Operations

List and Inspect (Read-Only)

# List applications
aws deploy list-applications --output table

# Get application details
aws deploy get-application --application-name <app-name>

# List deployment groups for an application
aws deploy list-deployment-groups --application-name <app-name> --output table

# Get deployment group details
aws deploy get-deployment-group \
  --application-name <app-name> \
  --deployment-group-name <group-name>

# List deployments
aws deploy list-deployments \
  --application-name <app-name> \
  --deployment-group-name <group-name> \
  --query 'deployments[:10]' --output table

# Get deployment details
aws deploy get-deployment --deployment-id <deployment-id>

# Get deployment target status
aws deploy list-deployment-targets --deployment-id <deployment-id>

# Get target details (instance-level status)
aws deploy get-deployment-target \
  --deployment-id <deployment-id> \
  --target-id <target-id>

# List deployment configs
aws deploy list-deployment-configs --output table

# Get deployment config details
aws deploy get-deployment-config --deployment-config-name <config-name>

# List on-premises instances
aws deploy list-on-premises-instances --output table

Create Applications and Groups

⚠️ Cost note: CodeDeploy to EC2/Lambda is free. CodeDeploy to on-premises instances: $0.02 per instance update.

# Create an application (EC2/On-premises)
aws deploy create-application \
  --application-name <app-name> \
  --compute-platform Server

# Create an application (Lambda)
aws deploy create-application \
  --application-name <app-name> \
  --compute-platform Lambda

# Create an application (ECS)
aws deploy create-application \
  --application-name <app-name> \
  --compute-platform ECS

# Create a deployment group (EC2 with tags)
aws deploy create-deployment-group \
  --application-name <app-name> \
  --deployment-group-name <group-name> \
  --service-role-arn <role-arn> \
  --ec2-tag-filters Key=Environment,Value=production,Type=KEY_AND_VALUE \
  --deployment-config-name CodeDeployDefault.OneAtATime \
  --auto-rollback-configuration enabled=true,events=DEPLOYMENT_FAILURE

# Create a deployment group with Auto Scaling
aws deploy create-deployment-group \
  --application-name <app-name> \
  --deployment-group-name <group-name> \
  --service-role-arn <role-arn> \
  --auto-scaling-groups <asg-name> \
  --deployment-style deploymentType=IN_PLACE,deploymentOption=WITH_TRAFFIC_CONTROL \
  --load-balancer-info targetGroupInfoList=[{name=<target-group-name>}]

# Create a custom deployment config
aws deploy create-deployment-config \
  --deployment-config-name MyConfig \
  --minimum-healthy-hosts type=FLEET_PERCENT,value=75

Deploy

# Create a deployment from S3
aws deploy create-deployment \
  --application-name <app-name> \
  --deployment-group-name <group-name> \
  --s3-location bucket=<bucket>,key=<key>,bundleType=zip

# Create a deployment from GitHub
aws deploy create-deployment \
  --application-name <app-name> \
  --deployment-group-name <group-name> \
  --github-location repository=<owner/repo>,commitId=<commit-sha>

# Create a deployment with description
aws deploy create-deployment \
  --application-name <app-name> \
  --deployment-group-name <group-name> \
  --s3-location bucket=<bucket>,key=<key>,bundleType=zip \
  --description "Release v1.2.3" \
  --file-exists-behavior OVERWRITE

# Register a revision
aws deploy register-application-revision \
  --application-name <app-name> \
  --s3-location bucket=<bucket>,key=<key>,bundleType=zip \
  --description "Version 1.2.3"

# Stop a deployment
aws deploy stop-deployment --deployment-id <deployment-id>

# Continue a blue/green deployment (switch traffic)
aws deploy continue-deployment --deployment-id <deployment-id>

Rollback

# Rollback is automatic if auto-rollback is configured
# Manual rollback: redeploy a previous revision
aws deploy create-deployment \
  --application-name <app-name> \
  --deployment-group-name <group-name> \
  --revision revisionType=S3,s3Location='{bucket=<bucket>,key=<previous-version-key>,bundleType=zip}' \
  --description "Rollback to previous version"

Delete / Destructive

🛑 DESTRUCTIVE — Always confirm with the user before executing any of these commands.

# Delete a deployment group
aws deploy delete-deployment-group \
  --application-name <app-name> \
  --deployment-group-name <group-name>

# Delete an application (deletes all deployment groups too)
aws deploy delete-application --application-name <app-name>

# Delete a deployment config
aws deploy delete-deployment-config --deployment-config-name <config-name>

Safety Rules

1. NEVER delete applications or deployment groups without explicit user confirmation.
2. NEVER expose or log AWS credentials, access keys, or secret keys.
3. ALWAYS confirm the application, deployment group, and revision before deploying.
4. ALWAYS recommend auto-rollback on deployment failure for production groups.
5. WARN about in-place vs blue/green deployment impact on running instances.
6. WARN that stopping a deployment may leave instances in an inconsistent state.

Best Practices

• Always enable auto-rollback on deployment failure.
• Use blue/green deployments for zero-downtime production deploys.
• Include a ValidateService lifecycle hook to verify the deployment succeeded.
• Use CodeDeployDefault.OneAtATime for production; AllAtOnce only for dev.
• Tag deployment groups for cost tracking and organization.

Common Patterns

Pattern: Monitor a Deployment

# Watch deployment progress
DEPLOY_ID="<deployment-id>"
while true; do
  STATUS=$(aws deploy get-deployment --deployment-id $DEPLOY_ID \
    --query 'deploymentInfo.status' --output text)
  echo "$(date): $STATUS"
  [ "$STATUS" = "Succeeded" ] || [ "$STATUS" = "Failed" ] || [ "$STATUS" = "Stopped" ] && break
  sleep 10
done

# Get per-instance status
aws deploy list-deployment-targets --deployment-id $DEPLOY_ID \
  --query 'targetIds' --output text | tr '\t' '\n' | while read tid; do
  aws deploy get-deployment-target --deployment-id $DEPLOY_ID --target-id "$tid" \
    --query 'deploymentTarget.instanceTarget.[targetId,status]' --output text
done

Pattern: Blue/Green with ECS

aws deploy create-deployment \
  --application-name <app-name> \
  --deployment-group-name <group-name> \
  --revision '{"revisionType": "AppSpecContent", "appSpecContent": {"content": "{\"version\": 0.0, \"Resources\": [{\"TargetService\": {\"Type\": \"AWS::ECS::Service\", \"Properties\": {\"TaskDefinition\": \"<task-def-arn>\", \"LoadBalancerInfo\": {\"ContainerName\": \"app\", \"ContainerPort\": 8080}}}}]}"}}'

Troubleshooting

Error	Cause	Fix
ApplicationDoesNotExistException	Application not found	Verify with list-applications
DeploymentGroupDoesNotExistException	Group not found	Check with list-deployment-groups
Deployment stuck in InProgress	Agent not running on instances	Check CodeDeploy agent status on target instances
HEALTH_CONSTRAINTS failure	Too many instances failed	Check minimum-healthy-hosts config; review instance logs
Lifecycle event timeout	Script exceeds timeout	Increase timeout in AppSpec or optimize script
Agent not found	CodeDeploy agent not installed	Install via aws ssm send-command or UserData