A fully static web application that catalogs alternative meanings for the acronym LGTM. Generate a random interpretation, browse the full collection, filter by category and rarity, and share individual entries via permanent URLs.

Live at: https://lgtm.airscript.it

LGTM collects hundreds alternative meanings for "LGTM" (Looks Good To Me), organized by:

• Rarity — common, regular, rare, legendary (weighted random generation)
• Category — funny, sarcastic, wholesome, dev, existential, corporate, chaotic

The site is fully static (no server-side rendering, no database). All data lives in data/lgtm.json.

Requires Node.js >= 24 and pnpm >= 10.

pnpm install

pnpm dev        # Start local dev server at http://localhost:4321
pnpm build      # Build static output to ./dist
pnpm preview    # Serve the built output locally

The navbar includes a 3-state theme toggle: system (default) → light → dark → system.

• Default follows the OS prefers-color-scheme media query
• The selected mode is persisted to localStorage under the key lgtm-theme
• A no-flash inline script in <head> reads the stored value and sets data-theme on <html> before CSS paints, preventing a flash of the wrong theme on reload

Uses Vitest with React Testing Library and jsdom.

pnpm test          # Run tests once
pnpm test:watch    # Run tests in watch mode

Test files live in src/test/ and cover:

• lgtm.test.ts — data loader functions and constants
• random.test.ts — weighted random selection logic
• ThemeSwitcher.test.tsx — 3-state cycle, localStorage persistence, aria labels
• RandomGenerator.test.tsx — rendering, navigation calls, generate interaction
• LegendaryEffect.test.tsx — particle component rendering and accessibility attributes

GitHub Actions runs on every push and pull request to main:

1. Checkout
2. Set up pnpm
3. Set up Node.js 24
4. pnpm install --frozen-lockfile
5. pnpm build
6. pnpm test

See .github/workflows/ci.yml.

Deployed as a static site on Vercel via the @astrojs/vercel adapter (static output mode). No server functions are used.

To deploy your own fork, connect the repository to a Vercel project. No environment variables are required.

1. Fork the repository
2. Add new entries to data/lgtm.json following the existing schema (numeric id, acronym: "LGTM", meaning, category, rarity, optional description and tags)
3. Keep IDs contiguous — the next ID is max(existing ids) + 1
4. Run pnpm build and pnpm test before opening a pull request