Field Table Planner is a small web application for boutique catering teams. It forecasts attendance, estimates food and beverage demand, sizes beverage packs, recommends staffing, and returns a price quote with risk notes.
python3.11 app.pyDefault port is 8000. If that port is already in use:
PORT=8011 python3.11 app.pyOpen http://127.0.0.1:8000 or the port you selected.
Two eval setups are available:
evals/contains the original YAML specs plus theplanner.eval_supporthelper layer.evals_v2/contains the wrapper-free experiment:vowel -druns the YAML-native surfaces directlypython3.11 evals_v2/run.pyruns the deeper class-method evals programmatically with serializers
Run the original eval set:
LOGFIRE_ENABLED=false LOGFIRE_IGNORE_NO_CONFIG=1 vowel -d /Users/mert/Desktop/test_dir/evals -v --hide-reportRun the wrapper-free CLI eval set:
LOGFIRE_ENABLED=false LOGFIRE_IGNORE_NO_CONFIG=1 vowel -d /Users/mert/Desktop/test_dir/evals_v2 -qRun the wrapper-free programmatic eval set:
python3.11 /Users/mert/Desktop/test_dir/evals_v2/run.pyThe interface is user-facing, but the core problem is operational:
- attendance forecasting
- consumption estimation
- inventory pack optimization
- staffing calculations
- pricing and risk scoring
Those utility-style functions exist because the project needs them, not the other way around.