Placeholder Academy

Master AI coding tools

A hands-on course that gets you fluent in the frontier-model coding tools developers actually use in the terminal — Claude Code (Anthropic), OpenAI Codex, and the Google Gemini CLI — plus how to choose between them and prompt each one well.

What you'll learn

Agentic coding tools run a frontier model directly in your terminal and editor. Instead of copying answers out of a chat window, they read your files, run commands, edit code, search the web, and carry out multi-step tasks on your behalf — asking permission at the points that matter. This course teaches the three leading tools side by side: what each does best, how their commands and workflows map onto one another, and when to reach for which.

Start free with the fundamentals and provider overviews. Upgrade to Premium for the full command reference, the five-module Advanced Track, and the provider deep-dives.

No accounts, no server. This site runs entirely in your browser. Your progress through each module — knowledge-check results, plus the XP, level, and badges you earn — is saved locally on this device (via localStorage). It isn't sent anywhere and won't follow you to a different computer.

Free vs Premium

The Free tier is a complete introduction: the Claude Code fundamentals, provider overviews for OpenAI and Google, a model-chooser guide, and the interactive playground. Premium unlocks the searchable Cheatsheet, the Advanced Track (5 modules), and the in-depth provider guides. See the Upgrade page for the full comparison and to enter a license key you already own.

The modules

Free — Fundamentals & overviews

Claude Clients

What Claude.ai, Claude Desktop, and Claude Code each do best, and when to reach for each one.

Core Concepts

Prompts, skills, the context window, tokens, and how to think about picking a model.

Getting Started

Installing Claude Code and running it for the first time.

Slash Commands & Skills

Built-in commands, custom commands, and packaged skills that extend what Claude Code can do.

Writing Effective Prompts

How to give Claude Code the context and structure it needs to get things right the first time.

Agentic Workflows & Subagents

Multi-step, tool-using workflows and how subagents help break down larger tasks.

OpenAI for Coding — Overview

The OpenAI coding surfaces — the agentic Codex CLI, the cloud, and the API — when to use each, and how they map onto Claude Code. A free taster for the deep-dive.

Google Gemini for Coding — Overview

The Gemini CLI, the Gemini API, and AI Studio — when to use each, the generous Google-login free tier, and how it compares to Claude Code and Codex.

Choosing a Frontier Model

A provider-neutral guide to picking among Claude, OpenAI/GPT, and Gemini — reasoning, speed, cost, context window, coding strength, and agent ecosystem, with a decision table.

Try It: Playground

Type a prompt and see illustrative, scripted responses from Claude.ai, Claude Desktop, and Claude Code side by side — a hands-on feel for how the surfaces differ.

Premium — Reference, Advanced Track & provider deep-dives

Claude Code Cheatsheet

A searchable quick reference for every Claude Code command — CLI commands and flags, /slash commands, bundled skills, and keyboard shortcuts — with copy-paste examples.

Advanced: Skills in Depth

SKILL.md anatomy, progressive disclosure, writing descriptions that auto-trigger, invocation control, and how custom commands are just skills.

Advanced: Loops & Goals

Drive bounded autonomy with /loop and /goal — fixed-interval vs self-paced runs, loop-until-done goals, and avoiding runaway cost with stop conditions.

Advanced: Workflows & Multi-Agent

Subagents via /agents, background and parallel agents, orchestration patterns (decompose, verify, judge panels), and when multi-agent beats a single agent.

Advanced: Automation, MCP & Hooks

Headless claude -p for CI, scheduled cron routines, plan mode, MCP servers, and settings.json hooks that automate on tool-use events.

Advanced: Context Mastery

CLAUDE.md memory scopes, output styles, checkpointing and /rewind, permission modes and sandboxing, and context hygiene with /clear, /compact, and /context.

OpenAI Codex CLI in Depth

Install & auth, interactive vs headless codex exec vs codex resume, slash commands, approval & sandbox modes, AGENTS.md, config, codex mcp, and codex cloud.

Gemini CLI in Depth

Install/channels, auth (OAuth vs GEMINI_API_KEY vs Vertex), interactive vs -p, flags, slash commands, GEMINI.md, MCP, and model selection.

Prompting Across Frontier Models

Portable prompt patterns, the instruction-file equivalents (CLAUDE.md / AGENTS.md / GEMINI.md), structured-output differences, and multi-model pipelines.

New here? Start with Claude Clients for the lay of the land, skim the free OpenAI and Gemini overviews, then use Choosing a Frontier Model to decide where to go deep. Ready for everything? Get Premium.