DEV Community

Cover image for Set up Cryptonym Desk on Sanity: build from a public dataset, no token
Christian Anderson
Christian Anderson

Posted on

Set up Cryptonym Desk on Sanity: build from a public dataset, no token

#javascript #tutorial #github #ai

The Sanity version of Cryptonym Desk keeps its word lists in a Sanity dataset instead of inside the page, with a proper review queue for new words. It now has a setup guide that GitHub runs weekly, with no Sanity token at all.

Five minutes, no account

The dataset is public-read, so building the site needs nothing but Node 20:

git clone https://github.com/casareanderson/cryptonym-desk-sanity && cd cryptonym-desk-sanity
npm ci && npm test        # 19 pass, 0 fail
npm run build             # 2 pages, straight from the live dataset
node scripts/proposals.mjs list
Enter fullscreen mode Exit fullscreen mode

The weekly check: tests, a build from the public dataset, the proposal queue, the corpus export

The check also exports the corpus with one GROQ query (104 documents: 20 digraphs, 78 word-bank entries, 6 presets) and builds the Studio. I wasn't sure the Studio would build without logging in. It does, so that's checked too.

Your own copy

The project id is written into the code rather than a setting, so the guide walks through it: create a project, export the public corpus with the same curl the CI uses, sanity dataset import it, and swap the id in a handful of files. Then the deploy workflow rebuilds your site on push and nightly.

The review queue app

Anything that needs a login (editing, moving proposals, deploying the review app) CI can't prove, and the guide says so. Full guide: docs/SETUP.md.

If you've moved a hard-coded list into a CMS, did it find bugs like it did here? Mine had a word in twice.

🤖 Drafted with AI assistance from my own homelab notes, logs and repos, then reviewed and edited before publishing.

Top comments (0)