observability-patterns

Configure Cloud Trace, BigQuery, AgentOps, Phoenix, and Weave for Google ADK agents.

Updated Aug 23, 2026
One-click install
npx skills add https://github.com/RawleySM/rlm-adk --skill observability-patterns-rawleysm
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: observability-patterns
Source: https://github.com/RawleySM/rlm-adk/tree/main/.gemini/skills/observability-patterns
Command: npx skills add https://github.com/RawleySM/rlm-adk --skill observability-patterns-rawleysm

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires gcloud, bq, gsutil, pip, curl, jq, python3, and includes scripts (resource) components.

What problem does it solve?

This skill provides templates and configurations to implement comprehensive observability in Google ADK agents, including logging, tracing, BigQuery analytics, and third-party observability tools (AgentOps, Phoenix, Weave). It helps you monitor agent behavior, debug performance, and set up tracing across tools when users mention observability, logging, tracing, BigQuery analytics, or related platforms.

Core Features & Use Cases

  • Cloud Trace integration templates for distributed tracing
  • BigQuery Agent Analytics plugin and data schemas
  • AgentOps session replays and telemetry
  • Phoenix/OpenInference instrumentation for self-hosted observability
  • Weave (W&B) tracing exporter for ML-oriented workflows

Quick Start

Run the provided setup scripts to enable a full observability stack for your Google ADK agents.

Frequently Asked Questions about observability-patterns

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

FAQPage Schema
How do I set up distributed tracing for Google ADK agents in production?

You enable distributed tracing for Google ADK by running the included setup scripts to wire Cloud Trace, sending agent operations as backend telemetry to monitor behavior and debug performance.

Can I use BigQuery for agent analytics and logging with ADK?

Yes, BigQuery analytics for ADK is supported via a dedicated plugin and data schemas, enabling detailed logging and querying of agent behavior directly within your Google Cloud data warehouse.

Does observability-patterns work with Phoenix and AgentOps for session replays?

Yes, the skill wires AgentOps for session replays and telemetry, while also providing Phoenix OpenInference instrumentation for self-hosted observability across your ADK agents.

What is the best way to monitor ML workflows in ADK using Weave?

For ML-oriented workflows, the skill provides a Weave (W&B) tracing exporter, allowing you to capture and monitor experimental agent traces alongside your existing machine learning operations.

Do I need gcloud and bq installed to enable multi-tool observability for ADK?

Yes, you need gcloud, bq, gsutil, python3, pip, curl, and jq installed, as the setup scripts rely on these dependencies to configure the full observability stack across Cloud Trace and BigQuery.

When should I not use a unified observability stack for agents?

A unified multi-tool observability stack may be unnecessary for simple local testing or non-production ADK agents, as the overhead of Cloud Trace, BigQuery, and third-party integrations targets production deployments.