Codex, Behind the Harness — Dominik Kundel, OpenAI
- Source
- AI Engineer
- Author
- AI Engineer
- Date
- token — The chunk of text a model reads and writes in — roughly three-quarters of a word — and the unit AI usage is billed in.
- inference — Running a trained model to get answers — the phase where AI is actually used, as opposed to trained.
- context window — The maximum amount of text a model can consider at once — its working memory for the current conversation or task.
- agent harness — The scaffolding around a model that turns it into a working agent — the loop, the tools it can call, and the rules for when to stop.
It is a rare first-party look at the engineering decisions inside a production coding , from transport to budgeting to approval design. Anyone building an agent harness can lift these tradeoffs directly.
“Once GPT 5.3 Codex Spark started serving a thousand tokens per second on Cerebras, inference stopped being the bottleneck and the network became it.”
“Tools can be marked deferred so they never enter the context window and surface through tool search when the model actually wants them, and the available skills list is capped at 2% of the context window, with descriptions trimmed as it grows past that.”
“Approval fatigue pushes people into full access, which their own security team hates, so an escalation now spins up an auto review subagent with read only permissions and no ability to spawn others, judging the action against the transcript and how explicitly the user authorized it.”
“Deleting a file you asked for reads differently from deleting a .git folder you never mentioned.”
videoI Gave an AI a Body — Cyrus Clarke, MIT Media Lab
videoThe Dark Arts of Skill Engineering — Paul Bakaus, Renaissance Geek
videoOperating Distributed Inference Systems at Scale — Nishant Gupta & Naman Ahuja, Meta
videoVertical Mobility: Inference from MVP to Trillion-Parameter Workloads — Sitanshu Gupta, CoreWeave
Checking sign-in…
Loading comments…