Containerization Assist: Docker and Kubernetes for an agent
containerization-assist
Azure MCP server takes an app from repo analysis to a Dockerfile, an image build, a vulnerability scan and Kubernetes manifests
- 45 ★ on the repository
- Last commit
- MIT
Broad access
Why this level
- The agent builds images through Docker and runs kubectl apply, which changes cluster state
- Deployment needs a working kubeconfig with deploy rights, and a manifest mistake can affect other workloads in the cluster
Install
- Works in
Manual install
claude mcp add containerization-assist -- npx -y containerization-assist-mcp startNeeds Node.js 20 and a running Docker.
Access and checks
How we reviewBroad accessThe tool can change data, run commands or handle secrets.
- The agent builds images through Docker and runs kubectl apply, which changes cluster state
- Deployment needs a working kubeconfig with deploy rights, and a manifest mistake can affect other workloads in the cluster
This is third-party code. Review the repository files before installing.
What it does
The containerization-assist-mcp server guides the agent through containerizing an app. It first analyzes the repository to detect language, framework and dependencies, then generates an optimized and hardened Dockerfile, builds the image, scans it with Trivy and suggests fixes for vulnerabilities. The image then gets tags, Kubernetes manifests are created, the cluster is prepared and kubectl apply runs with a health check. Version 3.0 added a policy system in Rego: strategy per environment (dev, staging, prod), cost limits by team tier, registry rules and consistency checks on the output. The same functions are available without MCP as a TypeScript SDK: analyzeRepo, buildImageContext and scanImage.
Who it is for. DevOps engineers and developers who need to package a service into a container and ship it to Kubernetes under team rules
Good fit when
- You need a Dockerfile for a new service without studying its stack by hand
- An image has to be scanned for vulnerabilities and a list of fixes returned
- An app needs Kubernetes manifests that follow team and environment policies
- You need uniform registry and resource rules across several projects
- You want to use the same functions in a VS Code extension without MCP through the SDK
Not a fit when
- There is no Docker: building and scanning need a running daemon and an accessible socket
- You only need ready Helm charts or Terraform: the server generates Dockerfiles and manifests
- No cluster is set up for deployment: a kubeconfig and deploy permissions are required
- You are after managing running containers: the catalog has separate Docker servers for that
Example request
Analyze this repository, write a Dockerfile, build the image, scan it for vulnerabilities and prepare Kubernetes manifests for dev.MCP
- Transport
- stdio
- Authentication
- not required
Environment variables
DOCKER_SOCKET- Path to the Docker socket, //./pipe/docker_engine on Windows
LOG_LEVEL- Log level, for example info
Limitations
Node.js 20 or newer and a running Docker are required; Trivy for scanning and Kubernetes with a working kubeconfig are optional. On Windows set DOCKER_SOCKET to the //./pipe/docker_engine pipe. Images are built on your machine through the local Docker, so large repositories and heavy base images take time and disk space, and scanning is unavailable without Trivy. The last commit was made on November 1, 2026.
How to disable. Remove the server from the client config, for example with claude mcp remove containerization-assist. Images already built and resources deployed to the cluster remain; remove them with docker rmi and kubectl delete.
FAQ
Do I need Kubernetes to use the server?
No. Analysis, the Dockerfile, build and image scan work with Docker alone; Kubernetes is needed only for manifests and deployment.
What scans the images?
Through Trivy, which is optional. Without it scanning is unavailable.
What is the policy system?
Rules in Rego that set defaults, filter the knowledge base and validate the output. Examples are in the repository.
Can I use it without MCP?
Yes, through the SDK: analyzeRepo, buildImageContext and scanImage are imported from containerization-assist-mcp/sdk.
Who publishes the server?
The Azure organization on GitHub; the project is MIT-licensed and listed in the MCP registry as io.github.Azure/containerization-assist.
Discussion
No questions or reviews yet.
Related
MCP serversMCP server for managing Docker in natural language: containers, images, networks and volumes, including remote hosts over SSH
maintenant monitoring MCP server: an agent reads container logs, check history and alerts but never touches the containers
CLI and app for workspaces: each AI agent gets its own container from a shared devcontainer.json, with an MCP server and a skill
MCP server for Kubernetes: the agent reads cluster state, diagnoses pod failures and runs operations through kubectl
Docker MCP by QuantGeekDev
docker-mcp
Docker MCP: an agent creates containers, deploys Docker Compose stacks from YAML, reads logs and lists containers. Runs through uvx for Claude Desktop
The docker mcp CLI plugin and gateway: MCP servers from Docker's catalog in isolated containers behind one connection point
Comparison with similar tools
Cards that share the main search query, and how they differ.
| Tool | Official | Authentication | Repository stars | Main difference |
|---|---|---|---|---|
| Containerization Assist: Docker and Kubernetes for an agent (this card) | yes | not required | 45 | |
| Docker MCP server | no | not required | 755 | manages containers and images through Docker, while this generates Dockerfiles and manifests from a repository analysis |
| Kubernetes MCP Server | no | not required | 2,244 | works with objects of an existing cluster, while this prepares the image and manifests before deployment |
README in short
The README describes an eight-step chain from repository analysis to deployment verification, requirements for Docker, Trivy and Kubernetes, manual VS Code setup and SDK examples. The Rego policy system and its examples for environments and teams have a separate section.