container-registry-setup
About
This Claude Skill helps developers set up and configure container registries like AWS ECR, Harbor, and Docker Hub. It provides guidance on registry selection, security, and CI/CD integration. Use it for implementing infrastructure-as-code, image scanning, and managing public or private repositories.
Quick Install
Claude Code
Recommended/plugin add https://github.com/majiayu000/claude-skill-registrygit clone https://github.com/majiayu000/claude-skill-registry.git ~/.claude/skills/container-registry-setupCopy and paste this command in Claude Code to install this skill
Documentation
Container Registry Setup Expert
Эксперт по настройке и управлению container registries.
Типы Registry
Cloud-Managed
| Registry | Provider | Features |
|---|---|---|
| ECR | AWS | IAM integration, scanning |
| Artifact Registry | GCP | Multi-format, regional |
| ACR | Azure | AD integration, geo-rep |
| Docker Hub | Docker | Public/private, CI/CD |
Self-Hosted
| Registry | Best For | Features |
|---|---|---|
| Harbor | Enterprise | RBAC, scanning, replication |
| Nexus | Multi-artifact | Maven, npm, Docker |
| Artifactory | Enterprise | Universal, HA |
| Distribution | Simple | Official Docker registry |
AWS ECR Setup
Terraform Configuration
resource "aws_ecr_repository" "app" {
name = "my-application"
image_tag_mutability = "IMMUTABLE"
image_scanning_configuration {
scan_on_push = true
}
encryption_configuration {
encryption_type = "KMS"
kms_key = aws_kms_key.ecr.arn
}
tags = {
Environment = "production"
Team = "platform"
}
}
resource "aws_ecr_lifecycle_policy" "cleanup" {
repository = aws_ecr_repository.app.name
policy = jsonencode({
rules = [
{
rulePriority = 1
description = "Keep last 10 images"
selection = {
tagStatus = "tagged"
tagPrefixList = ["v"]
countType = "imageCountMoreThan"
countNumber = 10
}
action = {
type = "expire"
}
},
{
rulePriority = 2
description = "Remove untagged after 7 days"
selection = {
tagStatus = "untagged"
countType = "sinceImagePushed"
countUnit = "days"
countNumber = 7
}
action = {
type = "expire"
}
}
]
})
}
ECR Authentication
# Login to ECR
aws ecr get-login-password --region us-east-1 | \
docker login --username AWS --password-stdin \
123456789.dkr.ecr.us-east-1.amazonaws.com
# Push image
docker tag myapp:latest 123456789.dkr.ecr.us-east-1.amazonaws.com/myapp:latest
docker push 123456789.dkr.ecr.us-east-1.amazonaws.com/myapp:latest
Harbor Self-Hosted
Docker Compose Setup
version: '3'
services:
harbor-core:
image: goharbor/harbor-core:v2.9.0
container_name: harbor-core
env_file:
- ./common/config/core/env
volumes:
- ./common/config/core/certificates:/etc/core/certificates
- ./common/config/core/key:/etc/core/key
depends_on:
- registry
- redis
- postgresql
networks:
- harbor
registry:
image: goharbor/registry-photon:v2.9.0
container_name: registry
volumes:
- registry_data:/storage
- ./common/config/registry:/etc/registry
networks:
- harbor
postgresql:
image: goharbor/harbor-db:v2.9.0
container_name: harbor-db
volumes:
- database:/var/lib/postgresql/data
environment:
POSTGRES_PASSWORD: ${DB_PASSWORD}
networks:
- harbor
redis:
image: goharbor/redis-photon:v2.9.0
container_name: harbor-redis
volumes:
- redis:/var/lib/redis
networks:
- harbor
nginx:
image: goharbor/nginx-photon:v2.9.0
container_name: nginx
ports:
- "80:8080"
- "443:8443"
volumes:
- ./common/config/nginx:/etc/nginx
depends_on:
- harbor-core
networks:
- harbor
volumes:
registry_data:
database:
redis:
networks:
harbor:
driver: bridge
Image Security
Vulnerability Scanning
# Trivy scan
trivy image myapp:latest
# Grype scan
grype myapp:latest
# ECR scan results
aws ecr describe-image-scan-findings \
--repository-name myapp \
--image-id imageTag=latest
Image Signing with Cosign
# Generate key pair
cosign generate-key-pair
# Sign image
cosign sign --key cosign.key myregistry/myapp:latest
# Verify signature
cosign verify --key cosign.pub myregistry/myapp:latest
Content Trust (Docker)
# Enable content trust
export DOCKER_CONTENT_TRUST=1
# Sign and push
docker push myregistry/myapp:latest
# Verify on pull
docker pull myregistry/myapp:latest
Kubernetes Integration
Image Pull Secret
apiVersion: v1
kind: Secret
metadata:
name: registry-credentials
namespace: default
type: kubernetes.io/dockerconfigjson
data:
.dockerconfigjson: |
eyJhdXRocyI6eyJteXJlZ2lzdHJ5LmNvbSI6eyJ1c2VybmFtZSI6InVzZXIi
LCJwYXNzd29yZCI6InBhc3MiLCJhdXRoIjoiZFhObGNqcHdZWE56In19fQ==
Deployment with Pull Secret
apiVersion: apps/v1
kind: Deployment
metadata:
name: myapp
spec:
replicas: 3
selector:
matchLabels:
app: myapp
template:
metadata:
labels:
app: myapp
spec:
containers:
- name: myapp
image: myregistry.com/myapp:v1.0.0
imagePullPolicy: Always
imagePullSecrets:
- name: registry-credentials
ServiceAccount Configuration
apiVersion: v1
kind: ServiceAccount
metadata:
name: myapp-sa
namespace: default
imagePullSecrets:
- name: registry-credentials
CI/CD Integration
GitHub Actions
name: Build and Push
on:
push:
branches: [main]
jobs:
build:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v4
- name: Set up Docker Buildx
uses: docker/setup-buildx-action@v3
- name: Login to ECR
uses: aws-actions/amazon-ecr-login@v2
- name: Login to Docker Hub
uses: docker/login-action@v3
with:
username: ${{ secrets.DOCKERHUB_USERNAME }}
password: ${{ secrets.DOCKERHUB_TOKEN }}
- name: Build and push
uses: docker/build-push-action@v5
with:
context: .
platforms: linux/amd64,linux/arm64
push: true
tags: |
123456789.dkr.ecr.us-east-1.amazonaws.com/myapp:${{ github.sha }}
myuser/myapp:${{ github.sha }}
cache-from: type=gha
cache-to: type=gha,mode=max
GitLab CI
stages:
- build
- push
variables:
IMAGE_NAME: $CI_REGISTRY_IMAGE
build:
stage: build
image: docker:24
services:
- docker:24-dind
before_script:
- docker login -u $CI_REGISTRY_USER -p $CI_REGISTRY_PASSWORD $CI_REGISTRY
script:
- docker build -t $IMAGE_NAME:$CI_COMMIT_SHA .
- docker push $IMAGE_NAME:$CI_COMMIT_SHA
only:
- main
Cleanup Scripts
ECR Cleanup
import boto3
from datetime import datetime, timedelta
def cleanup_untagged_images(repository: str, days_old: int = 7):
"""Remove untagged images older than specified days."""
ecr = boto3.client('ecr')
response = ecr.describe_images(
repositoryName=repository,
filter={'tagStatus': 'UNTAGGED'}
)
cutoff = datetime.now() - timedelta(days=days_old)
images_to_delete = []
for image in response['imageDetails']:
if image['imagePushedAt'].replace(tzinfo=None) < cutoff:
images_to_delete.append({
'imageDigest': image['imageDigest']
})
if images_to_delete:
ecr.batch_delete_image(
repositoryName=repository,
imageIds=images_to_delete
)
print(f"Deleted {len(images_to_delete)} images")
# Usage
cleanup_untagged_images('my-app', days_old=7)
Performance Optimization
Registry Caching
# Docker daemon.json
{
"registry-mirrors": ["https://mirror.gcr.io"],
"insecure-registries": [],
"max-concurrent-downloads": 10,
"max-concurrent-uploads": 5
}
Pull-Through Cache (Harbor)
# Harbor project config
replication:
- name: docker-hub-proxy
type: pull-through
source: https://registry-1.docker.io
filters:
- name: library/*
trigger:
type: manual
Troubleshooting
# Test connectivity
curl -v https://myregistry.com/v2/
# Check authentication
docker login myregistry.com
# Verify TLS
openssl s_client -connect myregistry.com:443 -servername myregistry.com
# Clear credentials
docker logout myregistry.com
rm ~/.docker/config.json
# Debug pull issues
docker pull myregistry.com/myapp:latest --debug
Лучшие практики
- Image immutability — используйте immutable tags
- Vulnerability scanning — scan on push обязателен
- Lifecycle policies — автоматическая очистка старых images
- Content trust — подписывайте production images
- Geo-replication — для global deployments
- Access control — минимальные права через RBAC
- Monitoring — алерты на failed pushes и pulls
GitHub Repository
Related Skills
sglang
MetaSGLang is a high-performance LLM serving framework that specializes in fast, structured generation for JSON, regex, and agentic workflows using its RadixAttention prefix caching. It delivers significantly faster inference, especially for tasks with repeated prefixes, making it ideal for complex, structured outputs and multi-turn conversations. Choose SGLang over alternatives like vLLM when you need constrained decoding or are building applications with extensive prefix sharing.
evaluating-llms-harness
TestingThis Claude Skill runs the lm-evaluation-harness to benchmark LLMs across 60+ standardized academic tasks like MMLU and GSM8K. It's designed for developers to compare model quality, track training progress, or report academic results. The tool supports various backends including HuggingFace and vLLM models.
langchain
MetaLangChain is a framework for building LLM applications using agents, chains, and RAG pipelines. It supports multiple LLM providers, offers 500+ integrations, and includes features like tool calling and memory management. Use it for rapid prototyping and deploying production systems like chatbots, autonomous agents, and question-answering services.
llamaguard
OtherLlamaGuard is Meta's 7-8B parameter model for moderating LLM inputs and outputs across six safety categories like violence and hate speech. It offers 94-95% accuracy and can be deployed using vLLM, Hugging Face, or Amazon SageMaker. Use this skill to easily integrate content filtering and safety guardrails into your AI applications.
