home / skills / physics91 / claude-vibe / docker-reviewer

docker-reviewer skill

/skills/docker-reviewer

This skill reviews Dockerfiles and docker-compose for optimization, security, and best practices to improve build speed and image safety.

npx playbooks add skill physics91/claude-vibe --skill docker-reviewer

Review the files below or copy the command above to add this skill to your agents.

Files (1)
SKILL.md
7.4 KB
---
name: docker-reviewer
description: |
  WHEN: Dockerfile review, multi-stage builds, layer optimization, docker-compose
  WHAT: Image optimization + Layer caching + Security scanning + Compose best practices + Build efficiency
  WHEN NOT: Kubernetes → k8s-reviewer, Terraform → terraform-reviewer
---

# Docker Reviewer Skill

## Purpose
Reviews Dockerfiles and docker-compose configurations for optimization, security, and best practices.

## When to Use
- Dockerfile code review
- Docker image optimization
- docker-compose.yml review
- Container security audit
- Build time optimization

## Project Detection
- `Dockerfile` in project
- `docker-compose.yml` or `docker-compose.yaml`
- `.dockerignore` file
- `Dockerfile.*` variants

## Workflow

### Step 1: Analyze Project
```
**Base Image**: node:20-alpine
**Build Type**: Multi-stage
**Compose**: v3.8
**Registry**: Docker Hub / ECR / GCR
```

### Step 2: Select Review Areas
**AskUserQuestion:**
```
"Which areas to review?"
Options:
- Full Docker review (recommended)
- Dockerfile optimization
- Layer caching strategy
- Security hardening
- docker-compose review
multiSelect: true
```

## Detection Rules

### Image Optimization
| Check | Recommendation | Severity |
|-------|----------------|----------|
| Large base image | Use alpine/slim/distroless | HIGH |
| No multi-stage build | Add build stage | MEDIUM |
| Too many layers | Combine RUN commands | MEDIUM |
| Installing dev deps | Separate build/runtime | HIGH |

```dockerfile
# BAD: Large image with dev dependencies
FROM node:20
WORKDIR /app
COPY . .
RUN npm install
RUN npm run build
CMD ["node", "dist/index.js"]
# Result: ~1GB image

# GOOD: Multi-stage with alpine
FROM node:20-alpine AS builder
WORKDIR /app
COPY package*.json ./
RUN npm ci
COPY . .
RUN npm run build

FROM node:20-alpine AS runner
WORKDIR /app
ENV NODE_ENV=production
COPY --from=builder /app/dist ./dist
COPY --from=builder /app/node_modules ./node_modules
USER node
CMD ["node", "dist/index.js"]
# Result: ~150MB image
```

### Layer Caching
| Check | Recommendation | Severity |
|-------|----------------|----------|
| COPY . before install | Copy package files first | HIGH |
| No .dockerignore | Add .dockerignore | MEDIUM |
| Changing files early | Order by change frequency | MEDIUM |

```dockerfile
# BAD: Cache invalidation on every code change
FROM node:20-alpine
WORKDIR /app
COPY . .                    # Invalidates cache on ANY change
RUN npm install             # Always reinstalls

# GOOD: Leverage layer caching
FROM node:20-alpine
WORKDIR /app
COPY package*.json ./       # Only invalidates on package change
RUN npm ci                  # Cached if packages unchanged
COPY . .                    # Code changes don't affect npm cache
RUN npm run build
```

```gitignore
# .dockerignore
node_modules
.git
.gitignore
*.md
.env*
dist
coverage
.nyc_output
```

### Security
| Check | Recommendation | Severity |
|-------|----------------|----------|
| Running as root | Add USER directive | CRITICAL |
| Latest tag | Pin specific version | HIGH |
| Secrets in build | Use build secrets | CRITICAL |
| No health check | Add HEALTHCHECK | MEDIUM |

```dockerfile
# BAD: Security issues
FROM node:latest           # Unpinned version
WORKDIR /app
COPY . .
ENV API_KEY=secret123      # Secret in image!
RUN npm install
CMD ["node", "index.js"]   # Running as root

# GOOD: Secure Dockerfile
FROM node:20.10-alpine AS builder
WORKDIR /app
COPY package*.json ./
RUN npm ci --only=production

FROM node:20.10-alpine
WORKDIR /app

# Create non-root user
RUN addgroup -g 1001 appgroup && \
    adduser -u 1001 -G appgroup -s /bin/sh -D appuser

COPY --from=builder --chown=appuser:appgroup /app/node_modules ./node_modules
COPY --chown=appuser:appgroup . .

USER appuser

HEALTHCHECK --interval=30s --timeout=3s --start-period=5s --retries=3 \
    CMD wget --quiet --tries=1 --spider http://localhost:3000/health || exit 1

EXPOSE 3000
CMD ["node", "index.js"]
```

### Build Secrets (Docker BuildKit)
```dockerfile
# syntax=docker/dockerfile:1.4

FROM node:20-alpine
WORKDIR /app

# Mount secret during build (not stored in layer)
RUN --mount=type=secret,id=npm_token \
    NPM_TOKEN=$(cat /run/secrets/npm_token) \
    npm ci

# Build command:
# DOCKER_BUILDKIT=1 docker build --secret id=npm_token,src=.npmrc .
```

### RUN Optimization
| Check | Recommendation | Severity |
|-------|----------------|----------|
| Multiple RUN for cleanup | Combine in single RUN | MEDIUM |
| No cleanup after install | Remove cache in same layer | MEDIUM |

```dockerfile
# BAD: Multiple layers, cache not cleaned
RUN apt-get update
RUN apt-get install -y curl
RUN apt-get clean

# GOOD: Single layer with cleanup
RUN apt-get update && \
    apt-get install -y --no-install-recommends curl && \
    apt-get clean && \
    rm -rf /var/lib/apt/lists/*
```

### Docker Compose
| Check | Recommendation | Severity |
|-------|----------------|----------|
| No resource limits | Add deploy.resources | HIGH |
| No health checks | Add healthcheck | MEDIUM |
| Hardcoded config | Use environment variables | MEDIUM |
| No restart policy | Add restart: unless-stopped | MEDIUM |

```yaml
# BAD: Minimal compose
version: '3.8'
services:
  app:
    build: .
    ports:
      - "3000:3000"
  db:
    image: postgres
    environment:
      POSTGRES_PASSWORD: password123

# GOOD: Production-ready compose
version: '3.8'

services:
  app:
    build:
      context: .
      dockerfile: Dockerfile
    ports:
      - "3000:3000"
    environment:
      - NODE_ENV=production
      - DATABASE_URL=postgresql://user:${DB_PASSWORD}@db:5432/app
    depends_on:
      db:
        condition: service_healthy
    restart: unless-stopped
    deploy:
      resources:
        limits:
          cpus: '1'
          memory: 512M
        reservations:
          cpus: '0.5'
          memory: 256M
    healthcheck:
      test: ["CMD", "wget", "-q", "--spider", "http://localhost:3000/health"]
      interval: 30s
      timeout: 10s
      retries: 3
      start_period: 40s

  db:
    image: postgres:15-alpine
    environment:
      POSTGRES_USER: user
      POSTGRES_PASSWORD: ${DB_PASSWORD}
      POSTGRES_DB: app
    volumes:
      - postgres_data:/var/lib/postgresql/data
    restart: unless-stopped
    healthcheck:
      test: ["CMD-SHELL", "pg_isready -U user -d app"]
      interval: 10s
      timeout: 5s
      retries: 5

volumes:
  postgres_data:
```

## Response Template
```
## Docker Review Results

**Project**: [name]
**Base Image**: node:20-alpine
**Build**: Multi-stage | **Compose**: v3.8

### Image Optimization
| Status | File | Issue |
|--------|------|-------|
| HIGH | Dockerfile | Using node:latest (~1GB) |

### Layer Caching
| Status | File | Issue |
|--------|------|-------|
| HIGH | Dockerfile:5 | COPY . before npm install |

### Security
| Status | File | Issue |
|--------|------|-------|
| CRITICAL | Dockerfile | Running as root user |

### Compose
| Status | File | Issue |
|--------|------|-------|
| HIGH | docker-compose.yml | No resource limits |

### Recommended Actions
1. [ ] Switch to node:20-alpine base image
2. [ ] Add multi-stage build
3. [ ] Add USER directive for non-root
4. [ ] Add resource limits in compose
```

## Best Practices
1. **Base Image**: Use alpine/slim/distroless
2. **Multi-stage**: Separate build and runtime
3. **Caching**: Order by change frequency
4. **Security**: Non-root, pinned versions, no secrets
5. **Compose**: Health checks, resource limits

## Integration
- `k8s-reviewer`: Kubernetes deployments
- `security-scanner`: Container security
- `ci-cd-reviewer`: Build pipelines

Overview

This skill reviews Dockerfiles and docker-compose configurations to improve image size, build speed, security, and runtime reliability. It highlights multi-stage build opportunities, layer-caching improvements, secret handling, and compose production hardening. The goal is actionable fixes you can apply in CI or locally to reduce image size, speed builds, and reduce attack surface.

How this skill works

The skill scans project files for Dockerfile variants, .dockerignore, and docker-compose YAML. It detects base-image choices, layer ordering that impacts cache, missing multi-stage builds, security issues (running as root, secrets in image, unpinned tags), and compose omissions (resource limits, healthchecks). It produces prioritized findings with concrete remediation steps and sample snippets for fixes.

When to use it

  • Perform code review on Dockerfiles before merge
  • Optimize image size and build time for CI pipelines
  • Audit container security and remove secrets from images
  • Harden docker-compose for production deployments
  • Detect layer-caching regressions after refactors

Best practices

  • Use slim/alpine/distroless base images and pin exact versions
  • Adopt multi-stage builds to separate build-time and runtime artifacts
  • Order COPY and RUN to maximize layer cache (copy package files before source)
  • Keep secrets out of layers; use BuildKit secrets or CI secret mounts
  • Add USER, HEALTHCHECK, resource limits, and restart policies in compose
  • Combine RUN steps and clean caches in the same layer to reduce image size

Example use cases

  • Convert a single-stage Node image into a multi-stage builder/runner to cut size
  • Reorder Dockerfile to COPY package*.json then npm ci to leverage cache
  • Add BuildKit secret mount to avoid embedding credentials during npm install
  • Add healthcheck, deploy.resources, and restart policy to docker-compose for production
  • Detect and suggest removal of large unwanted files via .dockerignore

FAQ

Will this skill fix files automatically?

No. It produces concrete recommendations and example patches; you apply changes in your repo or CI pipeline.

Does it scan Kubernetes manifests or Terraform?

No. It focuses on Dockerfiles and docker-compose. Use k8s-reviewer for Kubernetes and terraform-reviewer for Terraform.