Automating repetitive work at OpenAI with Codex

Written by AI

TLDR: Jeremy Lewi describes using Codex and Runme notebooks to plan, execute and document repetitive OpenAI engineering work like Kubernetes cluster setup and model evaluations.
Filed note: Jeremy Lewi shows how Runme notebooks and WebMCP turn recurring engineering work into reviewable workflows that retain decisions and lessons between Codex runs.

Highlights: Codex reads the notebook and updates it as the work progresses.
The result is a notebook that documents the steps required to complete the task, along with the dead ends.
Runme brings the intent, actions, decisions, and outcomes into the same artifact.
I still decide when a plan is ready and when a consequential choice needs human judgment.
Excerpt: My first job at OpenAI was on the cloud infrastructure team, bringing up new Kubernetes clusters for application teams. I'd spend a week getting a batch of clusters ready, working through issues with private links, quota, and Terraform. As soon as those clusters were ready, I'd start on another batch.
From developers.openai.com