Slim down your Docker images with the power of AI.

Container Diet is a futuristic, AI-powered CLI tool that analyzes your Docker images and Dockerfiles to provide actionable, "sassy but helpful" optimization advice. It helps you reduce image size, improve security, and follow best practices.

• 🧠 AI-Driven Analysis: Uses OpenAI (GPT-4o) to provide human-level, context-aware optimization tips.
• 🐳 Docker-Themed UI: Beautiful CLI output with Docker-brand colors and nautical icons.
• 🏠 Flexible Image Source: Analyze local daemon images, pull remote images with --remote, or auto-pull missing local images with --pull-missing.
• 🛡️ Security Focused: Detects root user violations, exposed secrets, and unnecessary packages.
• 🛠️ Auto-Fix: Automatically generate an optimized version of your Dockerfile with the --auto-fix flag.
• 🎭 "Container Dietician" Persona: Enjoy entertaining, roast-style feedback that keeps optimization fun.

• Go 1.21+
• Docker daemon running locally (required for local image analysis)
• Optional: Podman via Docker-compatible API socket
• OpenAI API Key

go install github.com/k1lgor/container-diet/cmd/cli@latest

git clone https://github.com/k1lgor/container-diet.git
cd container-diet
go mod tidy
go build -o container-diet cmd/cli/main.go

You must set your OpenAI API key before running the tool:

Linux/macOS:

export OPENAI_API_KEY="sk-..."

Windows (PowerShell):

$env:OPENAI_API_KEY="sk-..."

container-diet analyze my-app:latest

Providing the Dockerfile gives the AI more context for better suggestions.

./container-diet analyze my-app:latest --dockerfile Dockerfile

Use --remote to pull directly from a registry and analyze without requiring a local daemon image.

./container-diet analyze python:3.9-slim --remote

Use --pull-missing to keep local-first behavior, but auto-pull if the image is missing locally.

./container-diet analyze busybox --pull-missing

The most powerful feature! Use --auto-fix to have the Container Dietician write the optimized Dockerfile for you.

./container-diet analyze --dockerfile Dockerfile --auto-fix

This will generate a Dockerfile.diet file in the same directory. You can then compare it with your original and apply the improvements. Works even without a source Dockerfile by reverse-engineering the image layers!

For local analysis with Podman, expose Podman's Docker-compatible API socket and set DOCKER_HOST to it.

./container-diet analyze --help

Here is what happens when you feed the "Nightmare Monolith" Dockerfile to the Container Dietician:

Command:

./container-diet analyze --dockerfile samples/Dockerfile.nightmare

Output:

Reading Dockerfile: samples/Dockerfile.nightmare...

🐳 [AI ANALYSIS COMPLETE]
Asking the Container Dietician for insights... 🚢

Oh, honey, what do we have here? A "Monolith Monster" Dockerfile that's about to sink your ship
with its fatty layers and spicy security risks! Let’s roll up our sleeves and clean this galactic
mess. 🚀

---

⚠ WARNING: Version Drift Alert!
You've got a "fluffy" problem right at the start, darling! Using `ubuntu:latest` means you're
playing roulette with your build environment. 🎰

✓ SUGGESTION: Pin your base image to a specific version like `ubuntu:22.04` to keep things
predictable.

---

⚠ WARNING: Apt-get Avalanche!
Installing just about everything but the kitchen sink, are we? This is the definition of bloat,
my dear. 🐘

✓ SUGGESTION: Install only what's necessary for your app. Consider slimming it down by ditching
`openjdk-11-jdk`, `build-essential`, `cmake`, and `gdb` unless you truly need them. Otherwise,
you're just hoarding bytes.

---

⚠ WARNING: Hazardous Permissions Play!
777 permissions? I hope you’re wearing a hard hat! This is a security risk as wide as a black
hole. 🌌

✓ SUGGESTION: Use more restrictive permissions. Typically, `chmod -R 755` or `chmod -R 644`,
depending on what’s needed for executing.

---

⚠ WARNING: Root Cabal Alert!
Running SSH as root with root login permitted? You might as well hand over the keys to the
universe. 🔑

✓ SUGGESTION: Disable root login and use a non-root user. Also, ask yourself—do you really need
SSH in a container? Usually, it's a sign you need to rethink your strategy.

---

Oh, darling, let’s trim down that bloated ship before it swallows a sun. Your future workloads
will thank you for the speed and safety tunes! Now, get to work, 🛠️ and remember: less is always
more. 🐳✨

• cmd/cli: Main entry point for the CLI.
• internal/cli: CLI command definitions and logic.
• internal/ai: OpenAI client integration.
• internal/analyzer: Core image analysis logic (layers, size, config).
• samples/: Collection of Dockerfiles for testing (Light to Nightmare). See demo outputs.

MIT