Nucleos · Senior Software Engineer take-home

Khan Academy onboarding script and auto sign-in extension

An onboarding script creates one free adult-learner account. A Manifest V3 Chrome extension then signs that user in. Both report only what they can prove. If the login fails, you see Khan Academy's error text word for word. A success toast appears only after Khan Academy's own API confirms who is signed in.

Node + PlaywrightChrome MV3one permission: storage23 tests

Run it

npm run setup                                   # install + Playwright Chromium
npm test                                        # unit + end-to-end (offline)
npm run onboard -- --email you+nucleos@example.com
npm run use-account                             # give the verified account to the extension

Load the extension

1. Open chrome://extensions and turn on Developer mode.
2. Click Load unpacked and pick the unzipped folder (or extension/ in the repo).
3. Open khanacademy.org/login while signed out. The shipped config.json is deliberately wrong, so the first run shows Khan Academy's error.
4. Put the real account in config.local.json (git-ignored), reload the extension, and open the login page again to see the confirmed success toast.

How success is proven

• The account exists when Khan Academy's getFullUserProfile query returns a non-phantom user whose email is the one the script signed up with.
• You're signed in when the same check, run from the page you land on after login, returns the configured user. A URL change alone isn't proof, and anything less than a match gets an amber "not confirmed" toast.
• One attempt per visit, and a bounce back to the login page is reported, never retried. This keeps repeated failures from locking the account.

Full design notes, permissions, limitations and AI-tool disclosure are in the README.