onboard

name	onboard
description	First-run setup for the text-to-speech-toolkit plugin. Asks whether you target one TTS engine persistently or work across multiple, and saves the preference so other skills auto-prime.

Onboard

Initialize the text-to-speech-toolkit plugin by capturing your TTS engine preference and storing it persistently. Other skills in the toolkit will read this config to prime defaults and apply engine-specific markup automatically.

When to use

• First time using the text-to-speech-toolkit plugin
• You want to update your engine preference (re-runnable)
• You're setting up a new machine or workspace and need to establish your TTS defaults

Inputs to gather

• Your TTS engine preference: Do you work primarily with one engine, or several?
• If one: which engine (ElevenLabs, Google Cloud TTS, Azure Speech Services, AWS Polly, OpenAI TTS, generic SSML, or other)?

Procedure

1. Ensure the plugin data directory exists at ${CLAUDE_USER_DATA:-${XDG_DATA_HOME:-$HOME/.local/share}/claude-plugins}/text-to-speech-toolkit/. Create it if necessary.
2. Ask the user: "Do you primarily work with one TTS engine, or do you work across multiple engines?"
3. If one engine:
   • Offer a list: ElevenLabs, Google Cloud TTS, Azure Speech Services, AWS Polly, OpenAI TTS, generic SSML, other
   • Capture the choice
4. If multiple engines:
   • Note that the user will be asked per-skill which engine to target; no global default is set
5. Write config.json to the plugin data directory with this schema:

   {
     "default_engine": "elevenlabs" | "google" | "azure" | "polly" | "openai" | "ssml" | "multi",
     "set_at": "<ISO 8601 timestamp>"
   }
   

6. Read back the file and confirm the saved preference to the user.
7. Print the brand rule (non-destructive editing):

   Non-destructive editing: Every skill in this toolkit preserves your original files. When you run a skill, it creates original/ and edited/ subdirectories in your working directory. Your source file is never overwritten. You can always compare and choose which version to use.

Output / side effects

• config.json created (or updated) at ${CLAUDE_USER_DATA:-${XDG_DATA_HOME:-$HOME/.local/share}/claude-plugins}/text-to-speech-toolkit/config.json
• User sees confirmation of saved preference
• User is informed of the brand rule and non-destructive workflow

Safety / constraints

Re-runnable: This skill can be invoked at any time to update the engine preference. It overwrites the previous config.json gracefully.

Data directory creation: Ensure the full path to the plugin data directory is created if missing. Do not fail if the directory already exists.

No API validation: This skill does not validate that the user has API credentials or account access for the chosen engine. It only stores the preference for other skills to use.