Feature request: spoken responses (TTS with Claude's voice) in Claude Code

## Feature request

I use Claude Code in the desktop app and sometimes ask Claude to read status summaries aloud (e.g. while I'm away from the screen during long autonomous sessions).

Today the only option is piping text through macOS `say`, and the default system voices sound rough — especially for non-English languages (the compact Russian voice is barely listenable). Meanwhile the Claude mobile apps already have a great voice.

**Wish:** built-in text-to-speech output in Claude Code / desktop app — ideally the same voice as Claude's mobile voice mode. For example:
- a tool or setting that lets the agent speak short summaries/notifications aloud,
- useful for long-running sessions: "done, tests passed" spoken while you're across the room,
- multi-language support matters (my sessions are in Russian).

Thanks for the great product — filed at the user's request by Claude itself during a session. 🤖

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Start by examining the Claude Code desktop app entry points for output or notification settings and the existing macOS `say` workflow described in the issue. Compare the mobile app's voice behavior if accessible, then define a minimal spoken-summary path with Russian and other language support; done means short agent summaries or notifications can be spoken without piping text externally.

Written by the indexing model from the issue text.