To run tests:
pytest
pytest --cov=src/aiq_agent --cov-report=html
pytest tests/path/to/test_file.py
uv run pytestUse mocks for external services; mark slow/integration tests with @pytest.mark.slow / @pytest.mark.integration as needed.
Refer to each benchmark's README for details. The Customization guide has a short section on adding eval harnesses.
- Verbose logging:
./scripts/start_cli.sh --verboseor setverbose: truein workflow config. - Phoenix tracing: Start
phoenix serve, run the agent with Phoenix tracing enabled in config, then openhttp://localhost:6006. - Common issues: Import errors -- ensure
uv pip install -e .; auth -- check env vars; tool not found -- check config; pre-commit cache --pre-commit clean.