Installation
- pip
- uv
Setup
Call
instrument() once at application startup, before creating any ADK agents or runners.Usage
What gets traced
Token usage mapping
Thinking mode
When using Gemini models with thinking enabled, thought parts are automatically separated from answer parts. Thinking content is stored in metadata asreasoning_summary.
Example with tools
This example defines two function tools and runs anLlmAgent asynchronously using GoogleADKAdapter. We trace the agent via GoogleADKAdapter:
CHAIN (runner) → AGENT (city_info_agent) → LLM (model call) → TOOL (get_weather) → TOOL (get_population) → LLM (final response).
