Intermediate Artificial Intelligence 3.5 hours Featured

Claude Code Power User Guide: MCPs, Subagents, Hooks & Agentic Loops

Master every advanced Claude Code feature — from context engineering and CLAUDE.md files to MCP servers, custom subagents, hooks, plugins, and autonomous Ralph loops

2 Chapters

44 Lessons

253 min total

Free

Stop using Claude Code like a chatbot — become a genuine power user Most developers use Claude Code the way they once searched Stack Overflow: paste a problem, copy the answer, move on. This focused, fast-paced course is...

What you'll learn

Getting Started
Key Features & Efficient Usage

Engr. Mejba Ahmed

Course Instructor

Free · No card required

No payment — your email unlocks all 44 lessons

Be the first to review

Chapters 2

Lessons 44

Duration 3.5 hours

Level Intermediate

Certificate On Completion

Share this course

Enjoying this course?

Buy me a coffee

100% Free

44 lessons · certificate · no card

About This Course

Stop using Claude Code like a chatbot — become a genuine power user

Most developers use Claude Code the way they once searched Stack Overflow: paste a problem, copy the answer, move on. This focused, fast-paced course is for people who already use Claude Code and want to master its advanced features — the context engineering, automation, and extensibility that separate casual users from power users.

It's the deep end, on purpose. In a tight, high-density format you'll master:

Context engineering — sessions, compaction, and knowing when to start fresh; crafting great CLAUDE.md files and using Plan Mode effectively
MCP servers — connecting external tools and data, and managing permissions safely (including Docker and native sandboxing)
Custom subagents — creating and orchestrating specialist agents, and encouraging effective agent usage
Agent Skills — adding custom and third-party skills, and using skills as commands
Custom commands — building reusable prompt templates for your workflows
Hooks & plugins — automating pre/post tool-use events and extending Claude Code with plugins
Feedback loops — granting browser access, driving iteration with automated tests, and screenshot-based prompting
Autonomous operation — running Claude Code in a (Ralph) loop and using Claude Code Web in the cloud

Who this is for: developers already comfortable with the basics of Claude Code who want to squeeze real leverage out of it. By the end you'll drive Claude Code like a senior engineer drives their IDE — with custom subagents, skills, hooks, MCP servers, and automated loops doing the heavy lifting.

Who this course is for

Developers who know the basics and want production skills

Practitioners leveling up in Artificial Intelligence

Builders shipping real-world projects

Best if you're already comfortable with the basics and want production-grade depth.

Course Curriculum

2 chapters 44 lessons 253 min

6 lessons available to preview

1 Welcome To This Course!

4 Claude Code in Different Terminals

3min

5 Using Bun

3min

6 Base Usage & IDE Integration

Preview 8min

7 Configuring Claude Code

Preview 6min

8 Choosing AI Models

4min

9 Understanding Sessions & Context

5min

10 When To Start New Sessions & Making Sense Of Compaction

4min

11 Core Features You May Not Know Yet

5min

12 Advanced Permissions Management

7min

13 Running Claude Code via Docker Sandboxes

6min

14 Using Claude Code's Native Sandboxing

5min

15 Undoing Actions & The Importance of Version Control Systems

6min

16 Commands, Shortcuts & Settings Cheat Sheet

5min

17 Course Resources & Community

2min

1 Module Introduction

Preview 3min

2 Making Sense of Prompt & Context Engineering

4min

3 Prompt Engineering In Action & Working with Specs

10min

4 Prompt & Context Engineering Recommendation

4min

5 Initializing Claude Projects

5min

6 Crafting Great CLAUDE.MD Files

7min

7 CLAUDE.md vs "Auto Memory"

4min

8 Leveraging Plan Mode

7min

9 Using Claude Code's Built-in Tools

5min

10 Using MCP Servers & More On Permissions

8min

11 Understanding Subagents

6min

12 Creating & Using A Custom Subagent

7min

13 Encouraging Agent Usage

5min

14 Introducing Agent Skills

4min

15 Adding Custom Skills

12min

16 Using Agent Skills as Commands

4min

17 Enhancing Skills & Adding Third-Party Skills

5min

18 Iterating On The Demo App

9min

19 Building & Using Custom Commands (Prompt Templates)

8min

20 Using Screenshots For Prompting With Feedback

5min

21 Understanding & Using Hooks

11min

22 Installing & Using Plugins

6min

23 Creating Feedback Loops by Granting Browser Access

6min

24 Providing Feedback via Automated Tests

6min

25 Running Claude Code In A (Ralph) Loop

12min

26 Using Claude Code Web (Cloud)

8min

27 Course Wrap Up

3min

Your Instructor

Engr. Mejba Ahmed

AI Developer · Software Engineer · Entrepreneur

I build production AI systems and full-stack applications for a living, and I teach the exact workflows I use in real projects — not theory. Over 8+ years I've shipped 1,500+ projects, founded Ramlit Limited, and now build agentic AI tooling with Claude, GPT and open models. AI School is where I share that hands-on playbook so you can build and ship real work.

8+ years in production 1,500+ projects shipped Founder, Ramlit Limited