Quickstart Atlas

The front-end starter for the Callimacus quickstart: a natural-history site, animals and the places they live, built on an empty Callimacus project one step at a time.

Everything you need ships in this repository:

• data/ — 495 animals, 19 habitats, 493 knowledge documents, an illustration for every record, and skesis.config.json, the file that says which properties are searchable.
• blocks/ — one JSON file per block the guide adds, and blocks/pinned/ for part 4.
• config/ — the flow guidelines for part 3, the receptor for part 4, and the intent schema, use intent and landing for part 5.
• src/ — the real Atlas front end, already wired to your project through VITE_CALLIMACUS_CLIENT_ID.

Load the catalogue

Requires the Callimacus CLI 8.0 or newer, logged in against your project, on Node 26 or newer. Part 3 needs 8.0: thamyr config section set flow turns on the cognitive flow from config/flow.json. skesis up uploads the images the records name and swaps each {"$asset": "…"} for its url before submitting.

callimacus skesis init --dir data                        # confirms the committed configuration
callimacus skesis up --dir data                          # applies it, uploads the images, submits the records, waits until searchable
callimacus skesis status                                 # searchable (rev N)
callimacus document create --file data/knowledge.json    # the 493 documents, in batches of 50

skesis up is safe to re-run — records carry an id the engine matches on. document create is not: documents are inserted as submitted, so running that last line twice leaves two copies of all 493. Run it once against an empty project.

Run the front end

npm install
echo 'VITE_CALLIMACUS_CLIENT_ID=cal-pk-…' > .env   # your project's public Client ID
npm run dev

Then follow the guide from part 3: every block, receptor and agent setting is done in the Callimacus Studio or the CLI, and the page picks it up on the next question. No application code changes are needed.