Add Phonic Realtime Plugin

[qionghuang6]

Description

• This PR adds Basic Support for Phonic (https://docs.phonic.co/) as a Realtime Plugin.
• At this time, tool use and generateReply are not supported, but we plan on supporting these features in the future. Phonic does not support updateInstructions, updateChatCtx, and truncate
• Users can still use Webhook tools that they have defined on the Phonic platform (also executed on Phonic's platform) when passed in via phonicTools, or if included as a property of an agent defined on the Phonic platform (phonicAgent option)

Changes Made

• Add a new Phonic plugin, implementing RealtimeSession, mostly following the structure of the Gemini realtime plugin, with some simplifications.
• RealtimeSession uses the phonic SDK to connect to Phonic via WebSockets
• Although Phonic sends audio via the WebSocket at all times, we only use audio during assistant turns, aligning with the generations paradigm in the OpenAI and Gemini Realtime implementations.
• I have also not included anything to emit a usage event. The reason is because the RealtimeModelMetrics shape doesn't fit Phonic's billing structure, which is by number of seconds. I used the stt_metrics structure, but have removed it from this PR.

Demo

https://drive.google.com/file/u/3/d/1e3WdrlTozqGc7DG8eufa4qUVGSWkJZqF/view?usp=sharing

Pre-Review Checklist

• Build passes: All builds (lint, typecheck, tests) pass locally
• AI-generated code reviewed: Removed unnecessary comments and ensured code quality
• Changes explained: All changes are properly documented and justified above
• Scope appropriate: All changes relate to the PR title, or explanations provided for why they're included
• Video demo: A small video demo showing changes works as expected and did not break any existing functionality using Agent Playground (if applicable)

Testing

• Tested talking to the Phonic agent using the Livekit Agent Playground.
• Ensured that conversation items in the Livekit Agent Insights, on the Playground, and in ChatCtx are consistent with the recording on Phonic's end
• Ensured that config options are propagated properly.

Testing was done using the agent-starter-node repo, pointing to the plugin and agents-js locally:

    "@livekit/agents": "file:../livekit-agents-js/agents",
    "@livekit/agents-plugin-phonic": "file:../livekit-agents-js/plugins/phonic",

---

Note to reviewers: Please ensure the pre-review checklist is completed before starting your review.

[changeset-bot]

⚠️ No Changeset found

Latest commit: ff8a6b4

Merging this PR will not cause a version bump for any packages. If these changes should not result in a new version, you're good to go. If these changes should result in a version bump, you need to add a changeset.

This PR includes no changesets

When changesets are added to this PR, you'll see the packages that this PR includes changesets for and the associated semver types

Click here to learn what changesets are, and how to add one.

Click here if you're a maintainer who wants to add a changeset to this PR

[CLAassistant]

All committers have signed the CLA.

[toubatbrian]

Are these left empty on purpose or that they are not supported yet? Sould we add some kind of warning messages or throw if it is not supported?

[qionghuang6]

@toubatbrian commitAudio or clearAudio are not supported yet.
I left them empty because I saw that in your implementation of the Google realtime plugin at google/src/beta/realtime/realtime_api.ts:1583, these were left empty without logging a warning.
Would you prefer a logger warning here?

[qionghuang6]

Adding a warning log.

[toubatbrian]

Can we add some setup code sample in README? Also, it'd be good to showcase a example file under ./src/examples/ folder

[qionghuang6]

Sounds good, I'll add an example!

[toubatbrian]

Left some comments. Please also fix the LISCENCE issue in the CI, thanks

[qionghuang6]

I added an example and updated how instructions are used to better align with the other examples. Let me know if there's anything else I might want to look out for!