VoiceLabs: Managing Voice Library

Lists and manages the voice profiles on a VoiceLabs account: shows voices, adds built-in preset voices, suggests a voice for a language, and clones a voice from a recording only after the user confirms the speaker consented. Use when the user asks 'what voices do I have', 'add a built-in narrator voice', 'which voice fits French', 'clone my voice from this recording', 'create a custom voice', or wants to manage their voice library.

VoiceLabs voice library

Show the user their voices and add the ones they need. Built-in preset voices can be added freely; a cloned voice copies a real person and is handled with care.

Connecting and calling the tools

Every tool in this skill comes from the voicelabs MCP server, the VoiceLabs connector, and is written here as voicelabs:<tool>. The server lists the named tools and the Code Mode pair side by side.

• Tools listed by name: call them directly, for example voicelabs:list_voice_profiles.
• Several operations in one round trip, or a named tool is missing: call voicelabs:search_tools first, then call each tool inside voicelabs:execute_typescript as external_<tool> (for example external_list_voice_profiles); the program must return its result. Scopes, confirmations and gates are identical on both surfaces.
• "Tool not found": switch to the Code Mode route instead of retrying the direct call. If voicelabs:search_tools does not return the tool either, its OAuth scope was not granted; the user reconnects VoiceLabs and approves that scope.
• No voicelabs tools at all: VoiceLabs is not connected yet. Install the VoiceLabs power, which adds the server; it signs in with OAuth on first use.

Tools you will use

• voicelabs:list_voice_profiles: every voice on the account with id, name, description, language, voice_type (preset or cloned), default engine, and generation and sample counts. An empty account returns an emptyState message; relay it rather than reporting a failure.
• voicelabs:ensure_voice_profile: make sure a voice with a given name exists, creating it from a built-in preset (optional engine, builtinVoiceId, language, description). Idempotent: the same name returns the same profile with created: false.
• voicelabs:clone_voice_profile: create a cloned voice from a 2 to 30 second recording (name, referenceText with the exact words spoken, the audio as audioBase64 or audioUrl, optional language). Only present when the user granted the voice:clone scope.
• voicelabs:search_tools: lists the exact declarations, including the accepted engines and languages, when you need them.

Workflow

1. Call voicelabs:list_voice_profiles. Present the voices in a table: name, type, language, engine, generations.
2. To add a narrator or a voice in another language, call voicelabs:ensure_voice_profile with a clear name and the language. Report whether it was created or already existed, and its id for voicelabs:speak.
3. To clone a voice, first confirm the recording is of the user's own voice, or that the person speaking has agreed to have their voice cloned. Then check voicelabs:list_voice_profiles for a name clash, and call voicelabs:clone_voice_profile with the recording and the exact words spoken in it.
4. Offer the reading-text-aloud skill to try the voice.

Rules

• Never clone a voice without that confirmation of consent, and never clone a public figure or someone the user has not said agreed. If voicelabs:clone_voice_profile is not available, cloning happens in the VoiceLabs studio; say so.
• voicelabs:ensure_voice_profile only adds built-in presets; do not describe it as cloning.
• A duplicate name is refused on cloning; pick another name with the user.
• QUOTA_EXCEEDED on cloning means a plan limit on saved cloned voices; relay the message as written and stop.
• Deleting voices is not available through this connector; point the user to the studio.