devcc

devcc creates devcontainer templates for running AI coding agents in isolated, reproducible environments. Pick a language, pick an agent, and get a ready-to-use .devcontainer/.

Installation

pip install devcc

Quick Start

Step 1: See What languages and agents are available

devcc list

Languages:
  python               Python                    (default: latest)
  node                 TypeScript / Node.js      (default: latest)

Agents:
  claude-code          Claude Code CLI
  codex                Codex  CLI
  copilot              Copilot CLI
  cursor               Cursor CLI
  gemini               Gemini CLI

Step 2: Generate a devcontainer template

Go to your project repository and run:

cd /path/to/your/project
devcc create -l python -a claude-code

This creates a .devcontainer/ folder in your project:

.devcontainer/
├── devcontainer.json    # container configuration
├── system-setup.sh      # system setup (volume permissions, git-delta)
└── zsh-custom.sh        # shell customization (oh-my-zsh, autojump, aliases)

More Examples

# Pin a specific language version
devcc create -l python:3.11 -a claude-code

# Multiple languages in one container
devcc create -l python,node -a claude-code

# Multiple agents
devcc create -l python -a claude-code,codex

# Language only, no agent
devcc create -l node

# Custom output directory
devcc create -l python -a claude-code -o my-devcontainer

Step 3: Start the Dev Container

Option A: VS Code

1. Open your project in VS Code
2. Install the Dev Containers extension if you haven't already
3. Press F1 and select Dev Containers: Reopen in Container
4. VS Code builds the container and sets up your environment

Option B: CLI (using the Dev Container CLI)

# Install the CLI (once)
npm install -g @devcontainers/cli

# Build and start the container
devcontainer up --workspace-folder .

# Open a shell inside the container
devcontainer exec --workspace-folder . zsh

Once inside the container, your agent is ready to use (e.g., run claude for Claude Code).

What You Get

Every generated container includes:

• Ubuntu 24.04 base image
• Non-root user neo
• Workspace mount: your project is bind-mounted to /workspace
• Zsh and oh-my-zsh with a custom theme and plugins
• Your language(s) with the correct devcontainer feature and VS Code extension
• Your agent(s) installed automatically on container creation
• Persistent volumes: shell history and agent config survive container rebuilds

Validation

Validate devcontainer templates against the official devcontainer JSON schema:

devcc validate .devcontainer

You don't need to validate the devcontainer.json created by devcc. However, if you manually edit the generated devcontainer.json, you can run validation to ensure it remains compliant with the schema.