A practical introduction to AI agents for DevOps, SRE, and platform engineers. Covers what AI agents are at a code level (a Python example using OpenAI), how context windows work and their FIFO limitations, system prompts and the agents.md open standard, sub-agents with isolated context windows, and agent skills as on-demand prompts to keep context lean. Demonstrates spinning up a local Kubernetes cluster using OpenCode CLI with a custom Kubernetes expert sub-agent and skill files. Also briefly covers MCP for connecting agents to external systems like GitHub or a live Kubernetes cluster.
•22m watch time
411 Impressions1 Comment