My voices for Trillion
List, make, choose or remove the voices Trillion speaks to the person in, including their own voice cloned from a recording of them
Runs on the person's Mac (computer/speak.mjs speakTools.voices). Without input it lists every voice with the chosen one. make { recording, label?, choose? } makes the person's own voice from a recording of them in their files: the Mac hears it (mlx-whisper word timings) and keeps the stretch of about 9 seconds that ends where a sentence ends with a pause after it, with its exact words, as /Voice/<id>.wav, keeps it as a row of their Voices list, and chooses it unless choose is false. choose { voice id } makes Trillion speak in that voice; remove { voice id } removes one of the person's own voices. Every voice (the shared ones kept as rows the first time voices are listed), which one is chosen and every speed are the person's own list rows, also set on the Ambience screen and by any AI with put_list_rows; the reference recordings stay in their own files.
What it asks for
- make (Optional, details):
- choose (Optional, words): The voice id Trillion speaks in from now on
- remove (Optional, words): One of the person's own voice ids to remove
Use it from your AI
Connect the AI you already use to Trillion once. Then ask in your own words; your AI finds this tool by what it does and runs it.
Say to your AI: “Use Trillion to list, make, choose or remove the voices Trillion speaks to the person in, including their own voice cloned from a recording of them”
What your AI calls: run_tool {"tool_id":"tool:voices-on-this-mac"}