@vocea.app/mcp-server
Officialby vocea-admin
README.md
# @vocea.app/mcp-server
MCP (Model Context Protocol) server for Vocea — lets AI agents generate speech, transcribe audio, and manage voices.
## Installation
```bash
npm install -g @vocea.app/mcp-server
```
## Usage with Claude Desktop
Add to `claude_desktop_config.json`:
```json
{
"mcpServers": {
"vocea": {
"command": "vocea-mcp",
"env": {
"VOCEA_API_KEY": "vca_your_api_key"
}
}
}
}
```
## Available Tools
| Tool | Description |
|------|-------------|
| `vocea_generate_audio` | Convert text to speech |
| `vocea_list_voices` | List your cloned voices |
| `vocea_list_public_voices` | Browse community voices |
| `vocea_transcribe` | Transcribe audio (base64) to text |
| `vocea_get_credits` | Check credit balance |
| `vocea_get_audio` | Get audio details and CDN URL |
## Environment Variables
| Variable | Required | Description |
|----------|----------|-------------|
| `VOCEA_API_KEY` | Yes | API key starting with `vca_` |
| `VOCEA_BASE_URL` | No | Override base URL (default: `https://vocea.app/api`) |
## Running Locally
```bash
VOCEA_API_KEY=vca_xxx node dist/index.js
```
TDQS
A3.5/5.0
Scored across 6 tools
Disambiguation5/5
Each tool targets a distinct action: generating, retrieving, transcribing audio, listing different voice sets, and checking credits. No functional overlap exists.
Naming Consistency5/5
All tools use a consistent 'vocea_' prefix followed by a verb_noun snake_case pattern (e.g., generate_audio, list_public_voices). Even 'transcribe' fits as a single-word verb.
Tool Count5/5
Six tools cover the core TTS/STT domain without bloat. Each tool serves a clear, necessary purpose, making the set well-scoped.
Completeness4/5
Core workflows (generate, get, transcribe, check credits, browse voices) are covered. Missing features like voice cloning or deletion are minor gaps given the server's focus.