Add MiniMax text-to-speech processor

[octo-patch]

Reason: Add MiniMax text-to-speech processing so workflows can synthesize audio with a regional MiniMax account.

Add provider credentials and speech endpoint selection, an auto-discovered speech processor, and typed voice/audio controls. Support the configured speech models, optional pronunciation rules, voice effects and subtitles. Decode completed audio into an asset and reject API errors, incomplete responses, and invalid audio before creating an asset.

Validation:

• Seven unit tests passed, covering discovery, both regions, all supported models, request options, audio formats, and failure responses.
• Black 23.12.1, isort 5.13.2, and Flake8 7.0.0 passed on all changed Python files with the repository hook settings.
• Python compilation, git diff whitespace checks, and the configured secret scan passed.

The minimal provider configuration and registration allow this PR to run on main. Those shared registration files also appear in the existing chat PR #315 and will need reconciliation when both changes land.

[vercel]

@octo-patch is attempting to deploy a commit to the LLMStack Team on Vercel.

A member of the Team first needs to authorize it.