Your AI agent is deployed to Kubernetes, but external users still cannot reach it. Before learning how to solve this problem—controlling traffic flow, rate limiting abuse, terminating TLS, and autoscaling under load—you will own a traffic-engineer skill.
This skill becomes a component of your sellable Digital FTE portfolio. By the end of this chapter, you will have a production-tested skill that generates Gateway API configurations, rate limiting policies, and autoscaling rules for any AI deployment.
Output:
Before asking Claude to build your skill, define what you want to learn. Create a file named LEARNING-SPEC.md:
Output:
This specification tells Claude exactly what you need and how you will measure success.
Ask Claude to gather the authoritative source material:
Claude will retrieve documentation from the official Envoy Gateway project, giving your skill accurate, up-to-date patterns rather than hallucinated configurations.
Output:
Now prompt Claude to build the skill using the fetched documentation:
Claude will:
Output:
Your skill appears at .claude/skills/traffic-engineer/.
Verify your new skill generates valid Kubernetes YAML:
Output:
Save the output and validate:
Output:
If dry-run succeeds, your skill generates valid Gateway API configurations.
You now own a traffic-engineer skill built from official Envoy Gateway documentation. The rest of this chapter teaches you what it knows—and how to make it better.
Next: Lesson 1 — Ingress Fundamentals
Before moving on, consider: