SICSS 2026 — Synthetic Shoppers Challenge (participant materials)

Live mock shop:  https://sicss2026.derlem.com
Download kit:   https://sicss2026.derlem.com/downloads/sicss2026-participant-20260713.zip
Submit results:  https://sicss2026.derlem.com/submit
Leaderboard:     https://sicss2026.derlem.com/ranking

Files in this package:
  pyproject.toml                            — install CLIs (see docs/participant-code-guide.md)
  sicss2026/*.py                            — shop-vl-predict, shop-evaluate, optional CUA tools
  data/public/synthetic_shoppers_train.csv  — labeled training examples (90 rows)
  data/public/synthetic_shoppers_test.csv   — unlabeled test scenarios (38 rows)
  scenarios/R00xx.html                      — product pages (128 files, offline)
  screenshots/R00xx.png                     — product-page images (128 files)
  exercises/prompts/*.txt                   — starter persona prompts
  docs/sicss2026-task-description.md        — challenge instructions
  docs/participant-code-guide.md            — install + Python CLIs
  docs/challenge-api-contract.md            — language-agnostic API + BYOK guide

Install (from this folder, Python 3.11+):
  poetry install
  # or: python -m venv .venv && source .venv/bin/activate && pip install -e .

Open HTML locally: scenarios/R0001.html (or use the live site URLs in the CSV).

Run predictions (after install + your own API key):
  poetry run shop-vl-predict batch \
    --mode vision \
    --input data/public/synthetic_shoppers_test.csv \
    --output test_predictions.jsonl

Self-check on training labels:
  poetry run shop-evaluate -p my_train_preds.jsonl

You can submit your predictions (no account needed) at:
  https://sicss2026.derlem.com/submit
