# Audit a tool with evidence, not an ethics slogan: practice files

Original educational fixture, CC BY 4.0. All data is fictional or authored.

1. Read input.json and the rules in the handbook. Predict expected.json independently.
2. For manual practice, trace rows and calculations without a provider account.
3. For code practice, upload hosted-lab.ipynb to an authorised hosted Jupyter-compatible runtime. Install only requirements applicable to this fixture using its documented package interface, if not already present. Notebook access, compute and quotas are not guaranteed.
4. Alternatively place all files together in a permitted hosted Python workspace and run `python reference.py`.
5. Predict and perform all three changes in failure-tests.md; preserve evidence-log.csv.

Report error rates by explicitly supplied scenario group, disclose sample sizes and block release for unresolved critical privacy failures. These groups are test scenarios, not inferred sensitive identities.

For learners across regions underserved or excluded by mainstream providers: verify country eligibility, language, account/identity requirements, data terms, cost, quota and connectivity before choosing a cloud service. Use only legitimate available alternatives; never bypass restrictions or invent a country/account. No service access? Complete the labelled manual/fixture route without claiming a live AI run.

The reference makes no network calls, does not receive keys and never pretends to be a live provider. It checks this synthetic fixture only. Do not paste secrets into notebooks, downloaded work or public logs. The base result was checked during the platform build; no execution on an external learner service is claimed.