Feature request: read-aloud (TTS) for mobile Remote Control responses + screen-reader audit of the session view

Feature request

Add a read-aloud (text-to-speech) option for Claude's responses in the mobile app's Claude Code / Remote Control session view.

Use case

Operating a Claude Code session by phone via Remote Control is a natural fit for users who work by voice: low-vision users, users with limited reading stamina, and non-technical users for whom a spoken exchange is the whole interface. Input already works well through the OS keyboard dictation. Output has no spoken path inside the app.

Current state per the docs (checked 2026-08-21):

• voice-dictation.md: /voice is input-only, CLI-local, and "does not work in remote environments."
• mobile.md / remote-control.md: no mention of TTS or read-aloud for responses.
• accessibility.md: covers terminal screen-reader mode, magnifier, reduced motion, colorblind themes; nothing for the mobile surface.

Today the workaround is OS-level reading (iOS Speak Screen, Android Select to Speak), which works but reads raw screen content with no awareness of what is a response, a diff, or a permission prompt.

The accessibility half of the request

Separate from the TTS convenience: could the team confirm (or audit) that the Remote Control session view is fully navigable with VoiceOver/TalkBack: labeled controls, a traversable response stream, and reachable permission prompts? The terminal screen-reader mode in accessibility.md shows real investment in the CLI path for blind developers; the mobile session view deserves the same attention, and it is the surface a blind user is most likely to reach for when away from their rig.

Suggested shape

• A per-session or per-message "read aloud" action in the session view (system TTS voices are fine).
• Optionally, an auto-speak toggle for new responses while the session is in the foreground.
• A short accessibility note in mobile.md documenting whatever the supported path is.

Thanks — the mobile Remote Control feature is excellent, which is exactly why the audio gap stands out.