Astra and Fable as My Advisors, Human as the Judge, Codex and Claude Code as the Executors

Over the last six months, my working day has split cleanly into three layers: an advisor layer, a decision layer, and an execution layer. The advisor layer is OpenAI's GPT-6 Astra and Anthropic's Claude Fable 5.1; the decision layer is me, the Human; the execution layer is OpenAI Codex and Claude Code subagents, plus my own Hermes Agent skills. This post captures the September 2026 version of how those layers divide the work, which decisions I still keep on my side, how I hand plans across to the executors, and the loop I use to widen what the agents can run autonomously.