Summary

What you learned

In this lab you:

• Installed and configured the Okahu VS Code extension
• Connected to Okahu Cloud using GitHub authentication
• Configured and started the Okahu MCP Server
• Instrumented an AI agent with Monocle2AI
• Ran the agent and captured traces to Okahu Cloud
• Analyzed traces with nested spans in the Okahu extension
• Ran cloud evaluations (Hallucination, Sentiment) on agent traces
• Triaged and resolved issues with Okahu and a coding agent

Next steps

• CI/CD for Agentic AI

  ---

  Build a CI/CD pipeline that tests an AI agent, captures traces, and auto-assigns fixes.

  Start learning

• CI/CD for Non-Agentic Workloads

  ---

  Add zero-code instrumentation to an Azure provisioning pipeline with AI-powered root cause analysis.

  Start learning