What TypeSafe's Jev model is (a decision model, not an LLM) and how to use it with omp, Codex, Claude Code and Orca.
What an AI agent is, how the agent loop runs, and how context, harness and loop engineering grew around it.