Shortly AI

Shortly AI is a split production platform built around Supabase-backed agent runs.

• frontend/: Next.js dashboard for auth, run controls, and live status
• repo root: Python Playwright worker that claims queued runs and executes the agent
• platform/: supporting deployment, schema, and worker integration files

Production Targets

This repository is configured to support:

• Vercel for the dashboard
• Render for the dashboard and worker via [render.yaml](C:\Users\Tamada\Desktop\Shortly AI Agent\render.yaml)
• Railway for the worker via [railway.json](C:\Users\Tamada\Desktop\Shortly AI Agent\railway.json)

Deployment details live in [DEPLOYMENT.md](C:\Users\Tamada\Desktop\Shortly AI Agent\DEPLOYMENT.md).

Local Development

Install Python and frontend dependencies:

pip install -r requirements.txt
python -m playwright install chromium
npm install
npm --prefix frontend install

Copy [env.example](C:\Users\Tamada\Desktop\Shortly AI Agent\env.example) to .env and fill in your values.

Run the worker:

python worker.py

Run the dashboard:

npm run dev

Deployment Notes

• Recommended Vercel project roots: repository root or frontend
• Compatibility Vercel root: web
• Production worker default repo path: .
• Optional legacy worker override: AI_AGENT_REPO_PATH=external/ai-agent-browser

The web/ folder is a compatibility wrapper for older Vercel projects that were previously configured with web as the root directory. The real Next.js application still lives in frontend/.