/add-observability

Arguments: $ARGUMENTS

Goal: Add comprehensive observability capabilities (logging, tracing, analytics) to Google ADK agents for monitoring and debugging.

Core Principles:

• Detect existing agent configurations before modifying
• Understand observability requirements and scope
• Follow Google ADK best practices for instrumentation
• Integrate seamlessly with existing agent workflows

Phase 1: Discovery Goal: Understand project context and observability requirements

Actions:

• Parse $ARGUMENTS to determine observability type (logging, tracing, analytics, or all)
• Detect Google ADK agents in the project
• Example: !{bash find . -name "*.py" -type f | xargs grep -l "from google.adk"}
• Check for existing observability configurations
• Load package.json or pyproject.toml to understand dependencies

Phase 2: Analysis Goal: Understand existing agent structure and identify integration points

Actions:

• Read Google ADK agent files to understand current implementation
• Identify where observability hooks should be added
• Check for existing logging, tracing, or analytics libraries
• Determine which observability tools to use based on project setup

Phase 3: Requirements Clarification Goal: Gather specific observability needs from user

Actions:

• If $ARGUMENTS is unclear or incomplete, use AskUserQuestion to gather:
  • Which observability features are needed? (logging, tracing, analytics, or all)
  • What logging level should be configured? (DEBUG, INFO, WARNING, ERROR)
  • Should tracing include LLM call metadata?
  • What analytics platform to integrate with? (Google Cloud Logging, OpenTelemetry, custom)
  • Do you need real-time monitoring or batch analytics?

Phase 4: Implementation Goal: Integrate observability using specialized agent

Actions:

Task(description="Add observability to Google ADK agents", subagent_type="google-adk-observability-integrator", prompt="You are the google-adk-observability-integrator agent. Add observability capabilities to Google ADK agents based on $ARGUMENTS.

Context from Discovery:

• Observability type requested: $ARGUMENTS
• Detected Google ADK agents and configurations
• Existing project dependencies and structure

Requirements:

• Add appropriate logging instrumentation to agent lifecycle events
• Configure tracing for LLM calls and tool executions if requested
• Integrate analytics for performance metrics and usage tracking if requested
• Follow Google ADK patterns for middleware and hooks
• Ensure backward compatibility with existing agent code
• Add necessary dependencies to project configuration
• Create or update observability configuration files
• Document observability setup and usage

Expected output:

• Modified agent files with observability hooks
• Configuration files for logging/tracing/analytics
• Updated dependency files (package.json, pyproject.toml)
• Documentation on how to use and customize observability features")

Phase 5: Verification Goal: Ensure observability integration works correctly

Actions:

• Check that all required dependencies are installed
• Verify configuration files are properly formatted
• Example: !{bash python -m py_compile $(find . -name "*.py" -type f) 2>&1 | head -20}
• Review agent code for proper instrumentation
• Confirm logging/tracing/analytics outputs are being generated

Phase 6: Summary Goal: Report what was accomplished and next steps

Actions:

• Summarize observability features added
• List files modified and created
• Provide examples of how to view logs, traces, or analytics
• Suggest next steps:
  • Test observability in development environment
  • Configure production observability settings
  • Set up dashboards for monitoring
  • Review and adjust logging levels as needed