Scripts

The scripts/ directory holds developer tooling. It is not a Python package that ships in the runtime product; everything here is for contributors working in a repo checkout.

What This Directory Owns

• the unified test runner (run_tests.py) that drives the unit, runtime smoke, eval, and GUI functional suites
• UI and resource build helpers (build_ui.py, process_qss.py, compile_translations.py)
• developer install helpers (install.sh, install_helpers.sh). Native runtime sidecar (llama.cpp/whisper.cpp) builds live in Capsize-Games/airunner-native; install.sh downloads the pinned bundle from there.
• code-quality and reporting tooling (code_quality_report.py, coverage_report.py, the dead-code scanners and unused-import checkers, the complexity reports, mypy_shortcut.py, security_audit.sh)
• the out-of-band content-safety policy generator (build_policy_terms.py)
• local dev orchestration under scripts/dev/ (run_services.sh, run_gui.sh, test_services.sh, stop_services.sh)

The architecture audit and package map are tracked in docs/architecture/architecture-complexity-audit.md and docs/architecture/layered_product_architecture.md.

Usage

Run the unit suite from a repo checkout:

./venv/bin/python scripts/run_tests.py --unit

Rebuild the UI assets after changing .ui or .qrc sources:

./venv/bin/python scripts/build_ui.py

Regenerate the content-safety policy data file from a plaintext term list kept outside the repository (the default input lives under the gitignored tmp/ directory; never commit the plaintext list):

venv/bin/python scripts/build_policy_terms.py --input tmp/policy_terms_source.txt

See scripts/run_tests.py --help for the full set of test targets.