NVIDIA Physical AI Infrastructure for Codex: Check Readiness Gates
Plan NVIDIA Physical AI synthetic-data infrastructure by checking the order of cluster, serving and workflow readiness in a local configuration review.
中文Setup for Codex (NVIDIA plugin)
Codex with mock YAML or JSON status records. A real deployment uses the selected MicroK8s or Azure AKS environment, relevant NVIDIA components and operator access.
- Open the pinned GitHub source for this version. Copy the entire `skills/physical-ai-infrastructure-setup-and-resilient-scaling/` folder with its references into your scratch project's `.agents/skills/physical-ai-infrastructure-setup-and-resilient-scaling/`.
- Start Codex from that project and load the skill. Read the workflow's chosen target references after identifying the cluster, inference and OSMO responsibilities. Keep the exercise to local status evaluation.
Example task
Give Codex three local flags: clusterReady true, inferenceReady false, osmoReady true. Define an ordered readiness rule requiring all three for workload submission. Ask for the first failed stage and the permissible handoff state.Expected result
The serving stage is the first failed gate. Workload submission remains blocked in the fixture and the report directs attention to the inference status.
Upstream documentation
Original SKILL.md on GitHub · CC-BY-4.0 license · Source reviewed
Frequently asked question
Which Physical AI configuration should be checked first when readiness fails?
Check the first failing dependency in the deployment chain. A ready later component cannot satisfy an earlier prerequisite.