Beginner → Power user · 18 lessons · ~11 hours
How to Master Claude and ChatGPT: Everything in It
A complete operator's manual for the two frontier assistants. You will learn what the models actually are, walk through every shipped feature (Projects, Artifacts, Canvas, Memory, Code Interpreter, Custom GPTs, MCP, Claude Code), build a personal prompt system, and ship five real deliverables you can put in front of a client or a recruiter.
What you walk away with
- Explain tokens, context windows and reasoning modes in plain language
- Use every major feature of ChatGPT and Claude without guessing
- Write prompts with a repeatable 6-part structure instead of vibes
- Build a Custom GPT and a Claude Project for your own work
- Automate one real task end-to-end with the API (key placeholder provided)
Syllabus
1. Module 1 · The Machine Behind the Chat Box
What a large language model is, what a token is, why context windows matter, and why the model sometimes lies with total confidence.
- · What an LLM actually is (no maths yet)
- · Tokens, context windows and why long chats go stupid
- · Reasoning modes: when to pay for thinking
2. Module 2 · ChatGPT: Full Feature Walkthrough
Projects, Canvas, Memory, Code Interpreter, file & image input, voice, search, and Custom GPTs — what each one is for and the exact click path.
- · Projects, Memory and persistent context
- · Canvas and Code Interpreter: the two real power tools
- · Build your first Custom GPT
3. Module 3 · Claude: Full Feature Walkthrough
Artifacts, Projects, Styles, extended thinking, MCP connectors and Claude Code — Claude's model of work is different, and the difference is the point.
- · Artifacts: build apps inside the chat
- · Projects, Styles and extended thinking
- · Claude Code and MCP: the agent layer
4. Module 4 · The Prompt System (stop guessing)
A repeatable six-part prompt structure, few-shot examples, self-critique loops, and the failure patterns that waste your evenings.
- · The six-part prompt
- · Self-critique, rubrics and the second pass
- · Failure patterns and how to escape them
5. Module 5 · Workflows That Earn
Four end-to-end money workflows: research briefs, content systems, code review, and client deliverables — with the exact prompts.
- · The research brief workflow
- · Code review and debugging that actually works
- · Project: ship a paid-grade deliverable
6. Module 6 · APIs, Agents and Automation
Go beyond the chat box: the API shape, function calling, cost control, and a placeholder-key snippet ready for your own app.
- · The API in five minutes
- · Function calling, agents and when NOT to use them
- · Lab: wire a real endpoint