The Open Checklist Project is an open-data initiative to standardize and share trading card checklist data for collectors, developers, and the broader hobby ecosystem.
- Make reliable, structured checklist data freely available
- Define open schemas for consistent use across tools and apps
- Foster a contributor-friendly ecosystem for managing and improving trading card data
- ✅ Open YAML schemas — sets, manifests, and checklists (the v0.3 manifest form)
- 📦 Curated and contributed checklist data (organized by genre, sport, and set)
- 🛠 Tools for validating, transforming, and consuming data
- 🤝 GitHub workflows for contributor validation
The project uses standardized, independently-versioned schemas. As of v0.3, a set is
stored in the manifest form — a compact set.yaml + manifest.yaml + checklists/
from which the full per-card explosion is derived — replacing the older
one-file-per-card layout.
- Set v0.3 (manifest form):
schemas/set/schema.yaml(docs) - Manifest v0.2:
schemas/manifest/schema.yaml(docs) - Checklist v0.1:
schemas/checklist/schema.yaml(docs)
For the manifest form, the identity/UUID contract, versioning, validation, and examples, see the schemas directory.
To validate your data against the schemas, you can use Docker:
# Build the validation container
docker build -t open-checklist-validator .
# Run validation on all data files
docker run open-checklist-validatorAlternatively, if you have Python 3.11+ installed:
pip install pyyaml jsonschema
python tools/validate.pyPRs touching data/ also get an advisory computed review report; run it locally with
python tools/review_report.py data/<genre>/<set-id> (see CONTRIBUTING.md).
Open data is licensed under Creative Commons (TBD). Schema and code under MIT.