Skip to content

using-tools

Directory actions

More options

Directory actions

More options

Latest commit

 

History

History

using-tools

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
src
src
 
 
README.md
README.md
 
 
package.json
package.json
 
 

README.md

agent/using-tools

Voice agent with tool calling, using Deepgram STT and Cartesia TTS over WebSocket transport.

Demonstrates how to define tools in llmOptions.tools and handle tool calls via toolHook. Includes a get_weather tool that fetches real-time weather data from the Open-Meteo API.

Demonstrates:

  • Defining tool schemas in llmOptions.tools
  • Handling tool calls via the toolHook callback
  • Returning tool results with session.sendToolOutput()
  • Async tool execution (geocoding + weather API calls)
  • Selectable LLM vendor (OpenAI, Anthropic, Google, or Bedrock) via application variables

Setup

npm install

Application Variables

Configured in the jambonz portal and passed via session.data.env_vars:

Variable Required Default Description
LLM_MODEL No gpt-4.1-mini LLM model (OpenAI, Anthropic, Google, or Bedrock)
CARTESIA_VOICE No 9626c31c-... Cartesia voice ID
SYSTEM_PROMPT No (provided) System prompt for the voice agent

Environment Variables

Variable Default Description
PORT 3000 Port the WebSocket server listens on
LOG_LEVEL info Pino log level (debug, info, warn, error)

Running

npm start

Configure your jambonz application to use the WebSocket URL ws://your-server:3000/.

Note: Deepgram, Cartesia, and LLM provider credentials should be configured in the jambonz portal under speech provider settings. The Open-Meteo weather API is free and requires no authentication.