# Narova Elevenlabs

> Use this optional Narova companion when the user explicitly requests ElevenLabs, an ElevenLabs voice or voice ID, or hosted/premium cloud TTS for a Narova project. It installs and registers a separate ElevenLabs provider worker without adding cloud dependencies or credentials to the main Narova skill. Do not use it for ordinary local narrated-video requests; Narova remains local-first.

- Skill: `ammar-hasan/narova-elevenlabs` (Agent Skill, multi-file: 10 files)
- Install (CLI): `npx skillmds@latest add ammar-hasan/narova-elevenlabs`
- Raw SKILL.md: https://api.skillmd.com/api/skills/ammar-hasan/narova-elevenlabs/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: DevOps & Infra
- License: Apache-2.0
- Author: ammar-hasan (https://skillmd.com/u/ammar-hasan)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/ammar-hasan/narova-elevenlabs

---


# Narova + ElevenLabs

This skill is an optional external TTS provider for Narova. It owns all
ElevenLabs API details and speaks the generic `narova-tts-provider/v1` JSONL
protocol over stdin/stdout. Narova owns sentence caching and every downstream
audio/video stage.

## Setup

1. Install the standalone Narova CLI as described by the `narova` skill, then
   install `narova` and `narova-elevenlabs` as independently selected skills
   from `ammar-hasan/narova`.
2. Locate this installed skill from the directory containing this `SKILL.md`;
   call that absolute directory `<narova-elevenlabs-skill-dir>`.
3. Verify this worker:

   ```bash
   bash <narova-elevenlabs-skill-dir>/tool/setup.sh
   ```

4. Set the key in the environment used to run Narova. Never put it in
   `reel.config.mjs`:

   ```bash
   export ELEVENLABS_API_KEY="..."
   ```

5. Explicitly register this installed worker:

   ```bash
   narova providers add \
     <narova-elevenlabs-skill-dir>/tool/provider.json
   ```

6. Verify registration and authentication:

   ```bash
   narova providers doctor elevenlabs
   narova voices list --backend elevenlabs
   ```

Read [references/configuration.md](references/configuration.md) before writing
the voice block. It documents supported provider options, synthesis, errors,
security, billing, and removal.

When a brief needs a voice that does not exist yet, use
[references/voice-design.md](references/voice-design.md) and
`tool/design.py` to design previews from a description, audition them,
and create the permanent voice — the resulting voice ID slots into the
normal voice block.

## Operating rules

- Treat `speaker` as an ElevenLabs voice ID, not a display name.
- Keep `ELEVENLABS_API_KEY` only in the process environment or a user-managed
  secret store.
- Do not add ElevenLabs packages to Narova's venv or modify Narova's setup.
- Do not copy this worker into the Narova skill. Registration composes the two
  independently installed skills.
- Voice listing and synthesis are network operations and may be subject to
  account permissions, rate limits, and charges.
- Do not automatically retry synthesis failures. A response may have been
  billed even when the connection failed before Narova received it.
- Narova's sentence cache avoids billing again for unchanged sentences. A
  change to text, voice ID, language, tempo, gain, provider version, or
  `providerOptions` invalidates the relevant cache entry.