• static site (cloudflare pages)
• courses database (cloudflare d1)
• ai parse course from image (cloudflare workers)
• ai chat (cloudflare workers ai)

npm run dev, then localhost:300

or npm run build then push to git -> cloudflare ci/cd takes over

1. Copy .env.example to .env (or .env.local) and fill in:

   • CLOUDFLARE_ACCOUNT_ID – shown in the Cloudflare dashboard under Workers & Pages → Overview.
   • CLOUDFLARE_D1_DATABASE_ID – your D1 database id.
   • CLOUDFLARE_API_TOKEN – create via My Profile → API Tokens → Create Custom Token, granting “Cloudflare D1 → Edit” for your account.

2. Install deps if needed: npm install.

3. Dry-run the uploader:

   node adapter/upload.mjs uoft --session 20259 --code CSC258 --dry-run

4. Remove --dry-run when you’re satisfied to write into D1.

The Next.js app exposes GET /api/courses (and ?q=CSC111 prefix search) backed by Cloudflare D1.

• Required env vars at runtime:
  • CLOUDFLARE_ACCOUNT_ID
  • CLOUDFLARE_D1_DATABASE_ID
  • CLOUDFLARE_API_TOKEN

If you just want a local JSON snapshot (no Cloudflare credentials), use:

node adapter/export.mjs uoft --codes-file public/courses.txt

This writes:

• public/ttb-courses.json

used pretty heavily. refactor this component for me, write this component for me, other rudimentary code monkey tasks