La interfaz está localizada; las descripciones y documentos de las Skills se traducen automáticamente con IA a su idioma. El código, los comandos, rutas y términos técnicos se mantienen en el original, y prevalece el documento del autor.
Explorar Skills
Despliegue y operaciones · SKILL.md

adk-observability-guide

adk-observability-guide

ADK observability, Cloud Trace, prompt logging, agent analytics, BigQuery agent logs, ADK monitoring, ADK telemetry. Use when configuring tracing, logging, or analytics for ADK agents, debugging production agent behavior, or setting up monitoring for a deployed agent.

Independent GitHub snapshot · Automated classification · Original documentation governs

Documentación de la Skill

GitHub

name: adk-observability-guide description: "ADK observability, Cloud Trace, prompt logging, agent analytics, BigQuery agent logs, ADK monitoring, ADK telemetry. Use when configuring tracing, logging, or analytics for ADK agents, debugging production agent behavior, or setting up monitoring for a deployed agent." allowed-tools: Bash, Read, Write, Edit, WebFetch metadata: triggers: ADK observability, Cloud Trace, prompt logging, agent analytics, BigQuery agent logs, ADK monitoring, ADK telemetry related-skills: google-adk, agentic-ai-dev, python-dev domain: backend role: specialist scope: observability output-format: code last-reviewed: "2026-03-15"

ADK Observability Guide

Iron Law

NEVER ship a production ADK agent without Cloud Trace enabled. Set otel_to_cloud=True in the FastAPI app and LOGS_BUCKET_NAME env var before any production deployment.

Reference Files

FileContents
reference/cloud-trace-and-logging.mdCloud Trace setup, prompt-response logging infrastructure, environment variables, enabling/disabling locally, verification commands
reference/bigquery-agent-analytics.mdBigQuery Agent Analytics plugin — enabling, key features, tool provenance tracking
reference/slo-alerting.mdSLO-based burn-rate alerting (vs threshold alerting), multi-window pattern, error budget tracking, PromQL examples for availability/latency/safety SLOs

Observability Tiers

TierWhat It DoesDefault StateBest For
Tier 1: Cloud TraceDistributed tracing — execution flow, latency, errors via OpenTelemetry spansAlways-onDebugging latency, understanding agent execution flow
Tier 2: Prompt-Response LoggingGenAI interactions exported to GCS, BigQuery, and Cloud LoggingDisabled locally; enabled when deployedAuditing LLM interactions, compliance
Tier 3: BigQuery Agent AnalyticsStructured agent events (LLM calls, tool use, outcomes) to BigQueryOpt-in (--bq-analytics at scaffold time)Conversational analytics, custom dashboards, LLM-as-judge evals
Tier 4: SLO AlertingSLO-based burn-rate alerts (availability, latency, safety) via Prometheus — fires only when error budget is genuinely at riskOpt-in — requires Prometheus + Alertmanager stackProduction alerting for agentic AI services; replaces naive threshold alerts

Ask the user which tier(s) they need — they can be combined. Tier 1 is mandatory; Tiers 2, 3, and 4 are additive.


Quick Setup

Tier 1: Cloud Trace (Always-On)

# In your FastAPI app — scaffolded projects have this pre-configured
from google.adk.telemetry import setup_telemetry
setup_telemetry(otel_to_cloud=True)

View traces: Cloud Console > Trace > Trace explorer

Tier 2: Prompt-Response Logging

# Enable locally
export LOGS_BUCKET_NAME="your-bucket-name"
export OTEL_INSTRUMENTATION_GENAI_CAPTURE_MESSAGE_CONTENT="NO_CONTENT"

See reference/cloud-trace-and-logging.md for full infrastructure list, env vars, and verification commands.

Tier 3: BigQuery Agent Analytics

# Enable at scaffold time
uvx agent-starter-pack create . --bq-analytics

See reference/bigquery-agent-analytics.md for post-scaffold manual setup and key features.


Process

  1. Confirm tier(s) — Ask which tiers the user needs before configuring anything
  2. Read reference files — See reference/cloud-trace-and-logging.md (Tier 1 & 2) or reference/bigquery-agent-analytics.md (Tier 3) before writing any configuration
  3. Check env vars — Verify all required environment variables are set (table in reference/cloud-trace-and-logging.md)
  4. Verify with commands — Use the verification commands in reference/cloud-trace-and-logging.md to confirm telemetry is flowing
  5. Check costs — High telemetry volume? Switch to NO_CONTENT mode; reduce BigQuery retention; disable unused tiers

Troubleshooting

IssueSolution
No traces in Cloud TraceVerify otel_to_cloud=True in FastAPI app; check SA has cloudtrace.agent role
Prompt-response data missingCheck LOGS_BUCKET_NAME is set; verify SA has storage.objectCreator; check app logs
Privacy mode misconfiguredCheck OTEL_INSTRUMENTATION_GENAI_CAPTURE_MESSAGE_CONTENT — use NO_CONTENT for metadata-only
BigQuery Analytics not loggingVerify plugin in app/agent.py; check BQ_ANALYTICS_DATASET_ID env var
Traces missing tool spansTool spans appear under execute_tool — check trace explorer filters

Documentation Sources

TopicURL
Observability overviewhttps://google.github.io/adk-docs/observability/index.md
Agent activity logginghttps://google.github.io/adk-docs/observability/logging/index.md
Cloud Trace integrationhttps://google.github.io/adk-docs/integrations/cloud-trace/index.md
BigQuery Agent Analyticshttps://google.github.io/adk-docs/integrations/bigquery-agent-analytics/index.md
Skills guardadas