FlypBook: Scroll Like Shakespeare

See PLAN.md for the full plan and requirements.

Folders

• pipeline/generate_story.py: writes ONE story (plot plan → story → safety + logic check).
• pipeline/run_batch.py: the scheduled job. Handles queued requests first, then adds regular stories, and posts them to Supabase.
• supabase/schema.sql: database tables, security rules (Row Level Security), request queue, triggers.
• .github/workflows/generate-stories.yml: runs run_batch.py every 6 hours on GitHub Actions.

Secrets

Locally they go in pipeline/.env (see .env.example). Online they go in GitHub Secrets. Never commit them.

Name	Where to get it
GROQ_API_KEY	console.groq.com > API Keys
SUPABASE_URL	Supabase > Project Settings > Data API
SUPABASE_SECRET_KEY	Supabase > Project Settings > API Keys > Secret keys

Commands

cd pipeline
python generate_story.py                       # one story, saved locally only
python generate_story.py --request "..." --user sam
python run_batch.py                            # full job: requests + regular stories -> Supabase

Reviewing reported stories

Any report hides a story. In Supabase > Table Editor > stories, filter status = hidden, then set it to published (restore) or removed.

Changing limits

• Requests per user per day: Table Editor > app_settings > daily_request_limit.
• Stories per run and requests per run: top of pipeline/run_batch.py.
• How often it runs: cron line in .github/workflows/generate-stories.yml.