Subfork Examples

Data Aquarium

Data Aquarium example thumbnail

Upload an arbitrary UTF-8 CSV and turn up to 100 rows into a deterministic Canvas aquarium. Each fish represents one row. Choose a label field, then find the ten requested specimens by matching the requested value to the compact label rendered under each fish. Labels follow the selected field and truncate after 18 characters.

Try It

  1. Import the graph and upload a CSV in Upload aquarium CSV.
  2. Run the graph and open the recommended HTML Preview.
  3. Select any detected column as the fish label and play the expedition.
  4. Inspect the bounded parsed rows in the JSON panel.

CSV Parse reads the upload directly from Subfork artifact storage through an execution- and attempt-scoped grant with a 1 MB limit. It accepts UTF-8 CSV with an optional byte-order mark. The graph retains at most 100 rows and does not route the private artifact through HTTP Request. Gameplay, animation, scoring, and field selection run locally in the rendered page and make no network requests. Identical rows and field selection produce the same initial fish placement and mission order.

The rendered HTML contains the uploaded rows so the browser can play the game. Treat its preview/output as private when the CSV contains confidential or personal data, and do not publish it unintentionally. Labels and record details are inserted with DOM text APIs rather than interpreted as HTML.

There are no credentials, paid calls, sound, server-side scores, leaderboards, or persistent state. Empty CSVs and unusable columns show an error. Repository checks cover graph structure and script syntax; touch/keyboard accessibility and actual play after a fresh import remain manual checks.