Before releasing or assigning a module, verify:
- The engineering question, variables, units, assumptions, and validity limits are stated.
- Data provenance, rights, schema, and split rules are documented.
- A baseline case runs from a documented command in a clean environment.
- Dependencies and supported Python/framework versions are stated.
- Random seeds and model configuration are recorded where relevant.
- Outputs are traceable to code, data version, and parameters.
- Tests or smoke checks cover critical calculations and data schemas.
- Training, validation, and test data are not inadvertently mixed.
- Failure cases and known limitations are shown.
- Sensitive, student, confidential, or restricted material is excluded.