Run the Example

The repository ships one example, examples/basic-bridge - a complete, working embedding in about 20 lines. This page walks through it so you understand every moving part before writing your own.

What the example is

A single index.mjs that:

1. loads the env-driven config,
2. starts the bridge with startServer(),
3. opts into the SIGTERM/SIGINT drain (it owns the process).

That is the whole thing - and that is the point. Unlike the sibling bridges there are no tool or vision hooks to wire: your Line agent’s brain lives on Cartesia’s platform, and the bridge forwards the Teams context to it automatically (Your Line Agent).

Run it

git clone https://github.com/komaa-com/cartesia-msteams-bridge
cd cartesia-msteams-bridge/examples/basic-bridge
npm install
cp ../../.env.example .env    # fill in the three required values
npm start

The three values in .env:

Variable	What to put there
CARTESIA_API_KEY	Your Cartesia API key (server-side only; never rides the agent socket - each call mints a short-lived token).
CARTESIA_AGENT_ID	The deployed Line agent that should answer calls.
WORKER_SHARED_SECRET	The shared secret from StandIn pairing - both sides must match exactly.

Expose port 8080 with a tunnel (see Getting Started), set your StandIn identity’s agent WebSocket URL to the wss://…/voice/msteams/stream form, and place a Teams call - your Line agent answers.

The code

import { loadConfig, startServer } from "@komaa/cartesia-msteams-bridge";

const cfg = loadConfig();
startServer(cfg, undefined, { handleSignals: true });

console.log("basic-bridge example is up.");
console.log(`Point your StandIn identity's agent WebSocket URL at ws://<this-host>:${cfg.port}/voice/msteams/stream`);

• loadConfig() reads every setting from environment variables and fails loud on a missing required variable or a non-numeric number - a typo stops startup with a clear message instead of silently misbehaving.
• startServer(cfg, undefined, { handleSignals: true }) starts the WebSocket server. The second argument is the agent connector (tests inject a fake there; production uses the default). handleSignals: true wires SIGTERM/SIGINT to a graceful drain that ends every live call cleanly, lets an in-progress goodbye finish, and then exits the process - only enable it when the bridge owns the process, as this example does.

From example to your own service

• Embedding in a larger service? Leave handleSignals off and wire your own shutdown to the returned handle: await server.drain(); server.close(); - see Library API.
• Give the agent a deterministic opening line (CARTESIA_INTRODUCTION) and set the governor variables (MAX_CALL_MINUTES, CARTESIA_TTS_MODEL + CARTESIA_TTS_VOICE_ID, GOODBYE_TEXT) before going to production.
• Teach your Line agent code to consume the bridge’s call_context and goodbye_request events - shapes in Your Line Agent.
• For tests, inject a fake agent with the connectLine argument - see Library API.

If you only need the stock behavior, skip the embedding entirely and run the cartesia-msteams-bridge CLI.