Getting Started

Getting Started

Requirements: Python 3.11, Node.js 20+, SQLite (bundled). Postgres optional.

git clone https://github.com/tmdavid/Momboard && cd Momboard

python3.11 -m venv .venv
.venv/bin/python -m pip install -e '.[dev]'

cp .env.example .env      # set OPENAI_API_KEY, or see [[Local LLM Setup]]
mkdir -p data

.venv/bin/alembic upgrade head
.venv/bin/python -m app.seed                       # seeds the tag taxonomy
.venv/bin/python -m app.users create --email you@example.com --role admin

cd web && npm install && npm run build && cd ..
.venv/bin/uvicorn app.main:app --reload

Open http://localhost:8000, log in, hit ＋ New conversation, and paste a transcript — the fixtures in fixtures/ (see Fixtures and Evals) are made for exactly this first run.

The first five minutes

1. Upload fixtures/enforcement_heavy_user.vtt with any metadata.
2. Watch the pipeline chips: normalizing → tagging → analyzing → ready.
3. Open the conversation. Review suggestions keyboard-first: j/k to move, a accept, x reject.
4. Check the analysis sidebar — summary, top pains, commitments, and the Mom Test critique score.
5. Go to Explore, filter by ⚡ pain, and click Synthesize this view once you have a few conversations in.

Configuration

Everything is env-driven (.env): DATABASE_URL (SQLite default, swap to Postgres per DESIGN.md §8), LLM_BACKEND (openai | local), per-agent model names, SESSION_SECRET. Deployment (Docker, Fly.io, backups) is covered in the repo's DEPLOY.md.