Scaling Apps Beyond the Chat Window: Part 1, From Prompt to Project

What's inside

1. Level 1: Why One-Shot Chat Apps Stop WorkingFree
   Learn the four limits that stop a chat-built app from growing: the context window, no memory between chats, no history of changes, and nothing running online.
2. Level 2: What a Real Project Is: Files and FoldersFree
   Learn how a project is laid out as files and folders, what file extensions mean, and how to find a project folder on your computer.
3. Level 3: The Terminal Without Fear
   Open a terminal and use a small set of safe commands to move around and inspect a project folder.
4. Level 4: Choosing an Editor and a Coding Agent
   Understand what a code editor and a coding agent each do, compare the main kinds of tools fairly, and set up one of each.
5. Level 5: Packages and Running Your Chat App Locally
   Learn what packages and a package manager are, then move a chat-built app into a project folder and run it on your own computer.
6. Level 6: Plan Before You Prompt: The Instructions File
   Break a feature into small steps before asking the agent for code, and write a project instructions file that gives the agent lasting context.
7. Level 7: Describing Bugs and Reading Error Messages
   Read an error message to find what failed and where, and report bugs to the agent with expected and actual results and steps to reproduce.
8. Level 8: Reviewing AI Changes, Testing and Recovering
   Check what the agent changed, run simple tests to confirm the app still works, and recover when the agent breaks something.
9. Level 9: How Much Autonomy Should You Give AI? The Code Review Debate
   The live industry debate on whether to read AI-written code at all, set out fairly, and a practical way for a novice to decide how much to trust the AI on each change. Taught straight after the lesson on reviewing AI changes and before the Git lessons; refer to other lessons by name, never by number.
10. Level 10: What to Learn About Programming to Direct AI Well
    For people who are not developers: the small set of programming ideas that makes prompts meaningful and code readable, and a practical way to learn them alongside an AI assistant. Taught straight after the lesson on how much autonomy to give AI; refer to other lessons by name, never by number.
11. Level 11: Git Basics: Commits as Save Points
    Install Git, turn a project folder into a repository, and make commits as named save points.
12. Level 12: Branches as Safe Experiments and Going Back
    Use branches to try risky changes without touching working code, and use Git to undo changes or return to an earlier commit.
13. Level 13: GitHub: First Repository, Push and Pull
    Understand GitHub as the bridge off your device, create an account and a repository, and push and pull your project.
14. Level 14: A Clear README and a .gitignore File
    Write a README that explains the project and how to run it, and a .gitignore that keeps unwanted files out of the repository.
15. Level 15: Secrets, API Keys and .env Files
    Learn what API keys, secrets and tokens are, what must never be committed, and how to keep them in environment variables and .env files.
16. Level 16: When a Key Leaks: Rotate First, Then Clean
    Follow a fixed response when a secret reaches Git history or GitHub: revoke and rotate the key first, then remove it from the code and history.
17. Level 17: How an App Fits Together: The Map Ahead
    Understand the front end, back end, database and hosting, and how they connect, as the map for Parts 2 and 3.
18. Level 18: Picking a Stack AI Handles Well, Without Lock-In
    Choose a widely used, well-documented stack that coding agents handle well, keep it portable between tools, and check and date plans, limits and prices against official documentation.
19. Level 19: Final Challenge: Chat App to a Clean GitHub Repo Exam
    Bring the whole course together by turning a chat-built app into a clean, secret-free GitHub repository that runs locally and is ready to go online.

Access

The first 2 levels are free with a free account. Every level, the podcast edition and the AI tutor come with All Access at £4.99/month or any Creator plan — see pricing.

Start free →