# TTS integration notes

## Current state
- ElevenLabs API key is stored in local `.env`
- `.env` must never be committed

## Important note
The provided docs link was for Eleven Agents overview, not a plain text-to-speech request example.
For actual implementation, we need one of these:
- exact TTS endpoint docs link
- curl example that already works for this account
- minimal Python/JS sample that turns text into audio

## Expected input for v1
- source file: `examples/script_mistakes_when_learning_english_v2_3min.json`
- each item contains:
  - `start`
  - `duration`
  - `text`

## Expected output for v1
- single narration audio file
- preferred formats: mp3 or wav

## Next step
User provides a working ElevenLabs request example so the project can generate narration audio from the script.