Q&A

Off-topic-but-useful threads from the live session, kept here so they don't clutter the main walkthrough.

Claude project download

Claude chat has a download option in this demo, but it downloads a single tsx file rather than a full project.

That means the Claude chat app still needs a local shell around the file. Create the Vite project, paste the component, install dependencies, and run the app. For a full project export, the later bootstrapper category is a better fit.

Lovable with an existing GitHub repo

A participant asked whether Lovable can build an app from an existing GitHub repo. In this workshop, we cover the opposite direction. Create a prototype in Lovable, export it to GitHub, then continue in Cursor or VS Code.

Lovable is aimed at web development. If you want a Python backend, use Lovable for the frontend. Then build the backend separately with a coding assistant.

Score bug comment

A useful debugging suggestion is to check whether cleanup effects run correctly. In this app, the better check is to trace the score state and the game loop behavior directly.

Take the broader point beyond that specific bug. When an assistant makes a plausible suggestion, still test the behavior in the browser. Read the affected state updates before accepting the fix.

Recording link

We link the workshop recording from AI Coding Tools Compared: ChatGPT, Claude, Copilot, Cursor, Lovable and AI Agents, and that link covers the full technical walkthrough.