aAgentic AI AcademyLearn · Build · Ship

The Course

Master Claude Code, from first run to power user.

7 modules · 34 lessons · 7 hours. Start at the top, or jump to what you need.

—

Your progress

Loading your progress…

Start the course

What Claude Code Is (and Isn't)

1

Module 1Beginner

Getting Started with Claude Code

Go from nothing installed to a working, authenticated Claude Code session. Understand what makes Claude Code agentic (not autocomplete), the three-phase agentic loop, the surfaces it runs on, pricing tiers, and how to install across every platform and package manager. End the module able to start a session, steer it, and recover with checkpoints.

0/5

What Claude Code Is (and Isn't)

Agentic teammate, not an autocomplete

The Agentic Loop

Gather context, take action, verify results

Installing Across Every Platform

Native, Homebrew, WinGet, npm, apt/dnf/apk

Authentication, Plans & Access

Pro, Max, Team, Console, and CI tokens

Your First Session

Start, steer, interrupt, and rewind

2

Module 2Beginner

Core Workflows: Plan, Edit, Verify, Commit

Master the explore-plan-implement-commit workflow, permission modes, plan mode and adaptive thinking, verification-driven development, session and checkpoint management, and conversational steering. This is the muscle memory of productive Claude Code use.

0/5

Explore, Plan, Implement, Commit

The four-phase workflow that scales

Plan Mode & Adaptive Thinking

Read-only exploration before any edit

Permission Modes & Safety

Default, acceptEdits, plan, auto, bypass

Verification-Driven Development

Give Claude something to check itself against

Sessions, Checkpoints & Rewind

Resume, fork, branch, and time-travel

3

Module 3Intermediate

Memory & Context Engineering

Context is Claude Code's fundamental constraint. Learn the four-tier CLAUDE.md hierarchy, auto memory, path-scoped rules, imports, and the full context toolkit — /context, /compact, /clear, microcompaction, @-mentions, and how to keep long sessions productive without context rot.

0/4

CLAUDE.md & the Memory Hierarchy

Teach Claude your project's DNA

Auto Memory, Imports & Path-Scoped Rules

Notes Claude writes and rules that load on demand

Managing the Context Window

compact, clear, microcompact, and the 60% rule

@-Mentions, Multi-Dir & Image Hygiene

Pull the right things into context — and keep them out

4

Module 4Intermediate

Power Features: Subagents, Hooks, MCP & Slash Commands

Move from operator to power user: built-in and custom slash commands, subagents and parallel agents, deterministic hooks, MCP servers for external services, skills, plugins, and output styles. Learn when each mechanism is the right tool.

0/5

Built-in Slash Commands Tour

60+ commands across session, model, review, and workflow

Custom Slash Commands & Skills

Package your own workflows

Subagents & Parallel Agents

Isolated context for heavy, independent work

Hooks for Deterministic Automation

Guarantees CLAUDE.md can't make

MCP: Connecting External Services

Notion, Figma, databases, Slack, and custom tools

5

Module 5Advanced

Configuration, Automation & Scale

Take Claude Code to production: settings and permission rules, the headless print mode and Agent SDK, GitHub Actions and CI authentication, background and scheduled agents, git worktrees for parallelism, cost control, and IDE/desktop/web integrations.

0/5

Settings, Permission Rules & Config

settings.json precedence and the rules that govern Claude

Headless Mode & the Agent SDK

claude -p, structured output, and budgets

CI/CD, GitHub Actions & Background Agents

Claude Code in the pipeline and on the cloud

Worktrees, Parallelism & Cost Control

Scale quality and keep spend in check

IDE, Desktop, Web & Chrome Integrations

One engine, many surfaces

6

Module 6Advanced

Multi-Agent Orchestration at Scale

Go beyond a single subagent. This module covers Claude Code's four orchestration paradigms and when to reach for each: dynamic workflows that script tens-to-hundreds of subagents (up to 1,000 per run) with results kept out of your context window; experimental agent teams of coordinated peer sessions with a shared task list and inter-agent messaging; background agents and autonomous loops (/goal, /loop, agent view, routines) that work while you are away; and the ultra plan/review/ultracode modes that push planning and bug-hunting into the cloud at depth. You will learn the Workflow scripting API, the exact triggers and flags, the cost and permission model for each layer, and how to monitor, pause, and save these runs — plus the experimental-feature gotchas that bite in practice.

0/6

Choosing an Orchestration Pattern

Subagents, workflows, agent teams, and background agents

Dynamic Workflows: Scripted Fan-Out

One prompt to hundreds of subagents, results out of context

Agent Teams: Coordinated Peer Sessions

A lead, teammates, a shared task list, and a mailbox

Background Agents & Autonomous Loops

Work that runs while you're away: /goal, /loop, agent view, routines

Ultra Plan, Ultra Review & Ultracode

Cloud planning, multi-agent bug-hunting, and deep auto-orchestration

Cost, Observability & Orchestration Gotchas

Track the spend, attribute the agents, dodge the traps

7

Module 7Intermediate

Cheat Sheets, Shortcuts & Power Tips

A dedicated quick-reference module: the keyboard-shortcuts and CLI-flags cheat sheet, the complete slash-commands cheat sheet, a curated power-tips and best-practices lesson, and a patterns/anti-patterns/troubleshooting playbook. Built to be skimmed, bookmarked, and returned to.

0/4

Cheat Sheet: Keyboard Shortcuts & CLI Flags

The keys and flags you'll use every day

Cheat Sheet: Slash Commands

All 60+ commands, grouped and aliased

Power Tips & Best Practices

The habits that separate frustration from force-multiplier

Patterns, Anti-Patterns & Troubleshooting

Recognize the trap, apply the fix