AI-native coding where you describe intent and AI generates code
Claude Code CLI
Anthropic's terminal-based agentic coding tool
MCP
Model Context Protocol — standardized tool interface for AI
Context File
claude.md or agents.md — project instructions for AI
Checkpoint
Saved session state for rollback
YOLO Mode
Autonomous mode where AI runs commands without approval
Sub-Agent
Specialized agent within a larger swarm
Hook
Automated action triggered by events (pre-commit, post-deploy)
Swarm
Group of coordinated agents working in parallel
Sandbox
Isolated cloud environment for safe AI code execution
Cheatsheet
# Claude Code CLI Cheatsheet
# Install
npm install -g @anthropic-ai/claude-code
# Start session
claude
# Quick prompt
echo "explain this" | claude -p
# Session controls
/plan # Generate a plan
/review # Review changes
/summarize # Get session summary
/checkpoint # Save state
# Flags
--yolo # Autonomous mode
--model MODEL # Use specific Claude model
--resume # Resume last session
--verbose # Detailed output
# Config file (claude.json)
{
"mcpServers": { "serverName": { "command": "node", "args": ["server.js"] } }
}
# Project context (claude.md)
# Place in project root with conventions, stack, and key files.
Claude Model Comparison
Model
Best For
Context
Claude Haiku
Fast, lightweight tasks
200K
Claude Sonnet
Balanced speed/quality
200K
Claude Opus
Complex reasoning, code
200K
📘 Congratulations! You've completed the Claude Vibe Coding & Engineering course. You can now vibe-code prototypes, engineer production systems with Claude Code CLI, and orchestrate expert-level multi-agent swarms for complex applications.