Aria

Stack

• Python + Flask
• MongoDB
• Jinja (not EJS)
• Tailwind CSS + daisyUI

Run (from repo root)

npm install
npm run setup:py
npm run dev

Open http://127.0.0.1:5000

When to run each command

• npm install: only once (or when package.json changes)
• npm run setup:py: first time setup, and again only when requirements.txt changes
• npm run dev: every time you want to start the app locally

One command option

If you want setup + run together:

npm run dev:setup

This runs Python dependency install first, then starts the app.

What runs

• backend/run.py starts Flask routes
• frontend/templates/ contains Jinja templates
• frontend/static/css/output.css is generated automatically

Python dependencies installed by setup:py

• Flask
• pymongo
• python-dotenv
• and other backend packages from requirements.txt

Kokoro TTS Voice Setup (for all contributors)

Kokoro TTS (for demo voiceover) uses the af_bella voice by default, which is soothing and does NOT require espeak-ng for best quality.

• Python 3.10, 3.11, or 3.12 (Kokoro does NOT support 3.13+)
• kokoro (see requirements.txt)
• soundfile (see requirements.txt)
• misaki[en] (see requirements.txt)

Note: You do NOT need to install espeak-ng for the default af_bella voice. If you want to experiment with other voices, see the Kokoro docs for espeak-ng instructions.

Python version

Kokoro >=0.9.2 requires Python >=3.10 and <3.13. If you use Python 3.13, Kokoro will NOT install or work.

Troubleshooting

• If TTS sounds robotic or not like the demo, check that your Python version is correct and that misaki[en] is installed.
• If you get pip errors about kokoro version, check your Python version (must be <3.13).