adk-observability-guide

Configure OpenTelemetry tracing and prompt-response logging for ADK agents.

Updated Aug 27, 2026
One-click install
npx skills add https://github.com/SeanChenR/maestro-agent --skill adk-observability-guide
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: adk-observability-guide
Source: https://github.com/SeanChenR/maestro-agent/tree/main/.agents/skills/adk-observability-guide
Command: npx skills add https://github.com/SeanChenR/maestro-agent --skill adk-observability-guide

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

The ADK Observability Guide helps engineers and SREs configure, verify, and troubleshoot tracing, prompt-response logging, and analytics for ADK agents so they can understand execution flow, debug agent behavior, and audit GenAI interactions in production.

Core Features & Use Cases

  • Cloud Trace: Always-on distributed tracing via OpenTelemetry to visualize invocation span trees, latency, and errors across agents and tool executions.
  • Prompt-Response Logging: Configurable capture of GenAI interactions exported to GCS, BigQuery, and Cloud Logging for auditing, compliance, and replay.
  • BigQuery Agent Analytics: Optional plugin for structured event logging, SQL analytics, Looker Studio dashboards, and GCS offloading for multimodal content.
  • Third-Party Integrations & Troubleshooting: Guidance for AgentOps, Arize, Phoenix, MLflow, Monocle, Weave, and operational troubleshooting steps, IAM checks, and verification commands.

Quick Start

Enable scaffolded telemetry or set LOGS_BUCKET_NAME and OTEL_INSTRUMENTATION_GENAI_CAPTURE_MESSAGE_CONTENT then run the ADK service and verify traces and logs in the Cloud Console.

Frequently Asked Questions about adk-observability-guide

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I enable OpenTelemetry distributed tracing for ADK agents?

Enable distributed tracing for ADK agents by activating scaffolded telemetry or configuring OpenTelemetry environment variables, which visualizes invocation span trees, latency, and errors across agents and tool executions.

How do I log prompt-response interactions from GenAI agents to BigQuery and GCS?

Log prompt-response interactions by setting the LOGS_BUCKET_NAME and OTEL_INSTRUMENTATION_GENAI_CAPTURE_MESSAGE_CONTENT environment variables, exporting GenAI capture logs to GCS, BigQuery, and Cloud Logging for compliance auditing.

Does ADK observability work with deployments on Cloud Run, GKE, and Agent Engine?

Yes, ADK observability applies to scaffolded and manual deployments across Agent Engine, Cloud Run, GKE, and local development environments, supporting debugging and analytics workflows for all these platforms.

Can I use BigQuery for structured agent event analytics and dashboards?

You can enable an optional BigQuery analytics plugin at scaffold time to capture structured agent events, run SQL analytics, build Looker Studio dashboards, and offload multimodal content to GCS.

What is the best way to troubleshoot missing traces or logs in ADK agents?

Troubleshoot missing traces and logs by following operational steps, verifying IAM checks, and running verification commands in the Cloud Console to ensure telemetry and logging configurations are correctly applied.

Does ADK observability support third-party integrations like Arize and MLflow?

Yes, ADK observability provides guidance for integrating third-party tools including AgentOps, Arize, Phoenix, MLflow, Monocle, and Weave to enhance agent monitoring and troubleshooting workflows.