Codespaces

GitHub Codespaces Skill

Cloud development environments that work anywhere

GitHub Codespaces provides instant, configurable development environments in the cloud. This skill covers creating, managing, and optimizing Codespaces for any project type.

Prerequisites

• GitHub Account: With Codespaces access
• gh CLI: Install from https://cli.github.com/
• Authentication: Run gh auth login

Quick Reference

Core Commands

Machine Types

DevContainer Configuration

Location Options

.devcontainer/
├── devcontainer.json           # Default configuration
├── web-frontend/
│   └── devcontainer.json       # Frontend-specific
├── backend-api/
│   └── devcontainer.json       # Backend-specific
└── data-science/
    └── devcontainer.json       # Data science config

Basic devcontainer.json

{
  "name": "Project Dev Environment",
  "image": "mcr.microsoft.com/devcontainers/universal:2",
  "features": {
    "ghcr.io/devcontainers/features/node:1": {},
    "ghcr.io/devcontainers/features/python:1": {}
  },
  "customizations": {
    "vscode": {
      "extensions": [
        "GitHub.copilot",
        "ms-python.python"
      ]
    }
  },
  "postCreateCommand": "npm install",
  "forwardPorts": [3000, 5000]
}

Environment Profiles

This repository includes pre-configured profiles:

1. Standard (Default)

• Node.js + Python
• Common development tools
• GitHub CLI pre-configured
• Good for most projects

2. Web Frontend

• Node.js 20 LTS
• npm/yarn/pnpm
• Browser preview
• React/Vue/Angular extensions

3. Backend API

• Node.js + Python + Go
• PostgreSQL + Redis
• API testing tools
• Docker-in-Docker

4. Data Science

• Python 3.11+
• Jupyter notebooks
• pandas, numpy, scikit-learn
• Optional GPU support

Lifecycle Scripts

postCreateCommand

Runs once when container is created:

"postCreateCommand": "npm install && npm run setup"

postStartCommand

Runs every time codespace starts:

"postStartCommand": "npm run dev:services"

postAttachCommand

Runs when VS Code attaches:

"postAttachCommand": "echo 'Welcome! Run npm start to begin.'"

Secrets Management

Setting Secrets

# For a specific repo
gh secret set API_KEY --repo owner/repo

# For Codespaces specifically
gh secret set DB_PASSWORD --app codespaces

# For user (all Codespaces)
gh secret set NPM_TOKEN --user

Using Secrets in devcontainer.json

{
  "containerEnv": {
    "DATABASE_URL": "${localEnv:DATABASE_URL}"
  },
  "secrets": {
    "API_KEY": {
      "description": "API key for external service"
    }
  }
}

Port Forwarding

Automatic Forwarding

{
  "forwardPorts": [3000, 5000, 8080],
  "portsAttributes": {
    "3000": {
      "label": "Frontend",
      "onAutoForward": "openBrowser"
    },
    "5000": {
      "label": "API",
      "onAutoForward": "notify"
    }
  }
}

Manual Forwarding

# In codespace terminal
gh codespace ports forward 3000:3000 -c <codespace-name>

Performance Optimization

Prebuilds

Enable prebuilds for faster startup:

// .github/codespaces/prebuild.json
{
  "triggers": {
    "branches": ["main", "develop"],
    "paths": ["package.json", ".devcontainer/**"]
  }
}

Dotfiles Repository

Configure personal dotfiles:

# In GitHub Settings > Codespaces
# Set dotfiles repository to your-username/dotfiles

Resource Management

# Check machine type
gh codespace list --json name,machinetype

# Resize codespace
gh codespace edit -c <name> -m largePremiumLinux

Common Workflows

Quick Start New Feature

# Create codespace for feature branch
gh codespace create -r owner/repo -b feature-branch -m standardLinux

# Open in VS Code
gh codespace code

PR Review in Codespace

# Create codespace from PR
gh codespace create -r owner/repo --branch pr-branch

# Or use the PR directly
gh pr checkout 123
gh codespace create

Share Environment

# Export port with visibility
gh codespace ports visibility 3000:public -c <name>

# Get shareable URL
gh codespace ports -c <name>

Troubleshooting

Cost Management

Auto-stop Settings

Configure in GitHub Settings > Codespaces:

• Default idle timeout: 30 minutes
• Retention period: 7 days

Best Practices

1. Stop when not in use: gh codespace stop
2. Delete completed work: gh codespace delete
3. Use appropriate machine size: Don't over-provision
4. Enable prebuilds: Faster startup, less billing

Related Documentation

• Codespaces Workflow Guide
• Reference Documentation
• Codespaces Manager Agent
• DevContainer Configurations