Kitten sitting
A development example for two synthetic households, their kittens and their sitting bookings.
Each run owns a separate SQLite database. The adapter uses Python's standard library and the
addresses in the scenario use example.invalid.
Run from the company repository root:
python3 tools/testdata/testdata.py create \
--manifest tools/testdata/examples/kittens/testdata.json \
--scenario two-households --environment development \
--state-root /absolute/private/testdata
Use the returned run_id with populate, exercise, status, reset and delete:
python3 tools/testdata/testdata.py populate \
--state-root /absolute/private/testdata --run RUN_ID
python3 tools/testdata/testdata.py exercise \
--state-root /absolute/private/testdata --run RUN_ID
populate inserts two households, two kittens and two bookings. exercise reads each booking
through its household ID and captures one booking confirmation per household in SQLite.
Repeated calls keep those counts stable. reset restores the scenario and clears captured
confirmations. delete removes this run's database; the shared runner retains its receipt.
The database records its run ID and verifies ownership before reuse or deletion. It uses composite foreign keys to keep each booking's kitten in the same household. The lifecycle tests verify that another household's booking lookup returns an empty result, that a booking referencing another household's kitten fails, and that resetting or deleting one run preserves a second run. This example demonstrates local data and query isolation. Real login, network authorization and provider integrations belong to a product adapter's own tests.