Add text-to-speech using Web Speech API

[neubig]

This PR adds text-to-speech functionality to the chat interface using the Web Speech API. Features:

• Uses browser built-in Web Speech API for text-to-speech
• Only speaks assistant messages (not user messages)
• Can be toggled on/off with a button in the chat interface
• Automatically stops speaking when disabled
• Removes markdown formatting before speaking
• Tries to use a good English voice if available
• Cancels any ongoing speech when a new message arrives

The implementation is simple and works directly in the browser without additional API calls or dependencies.

All speech-related tests are passing:

• Chat message speech tests
• Speech slice tests
• Toggle speech button tests

---

To run this PR locally, use the following command:

docker run -it --rm   -p 3000:3000   -v /var/run/docker.sock:/var/run/docker.sock   --add-host host.docker.internal:host-gateway   -e SANDBOX_RUNTIME_CONTAINER_IMAGE=docker.all-hands.dev/all-hands-ai/runtime:e95ad75-nikolaik   --name openhands-app-e95ad75   docker.all-hands.dev/all-hands-ai/openhands:e95ad75

[neubig]

This is not super-high priority and I'm having trouble getting it to work, so I'll close this for now.