refactor: normalize prompt generation, implement robust section parsing, and update inference parameters for stability

chore: clean up run.sh — scope cleanup to project root, improve readability

refactor: split flat root into env/, core/, ui/ domain packages

fix: mount modal/ dir so dataset.py is importable in container, expand cache cleanup to repo root

feat: restructure remote training files under modal/ and expand synthetic CBT dataset

chore: remove redundant virtual environment activation calls from run.sh

chore: remove local pip cache configuration and add console logging during setup

chore: activate python virtual environment in run.sh

Fix UI styling: kicker padding, centering, and vertical formatting of list items in output cards

refactor: simplify project structure to flat root level layout, add pyright typechecking, and add code comments

feat: add sponsor model strategy selector

refactor: simplify project structure

docs: expand NeighborDocs hackathon scaffold

feat: scaffold NeighborDocs MVP