rhoai-project-scoped-resources

name	rhoai-project-scoped-resources
metadata	{"author":"rhoai3-demo","version":"1.0.0","platform-family":"rhoai","platform-baseline":"repo","ocp-baseline":"repo","skill-group":"RHOAI Platform"}
description	Use when documenting, reviewing, or rebuilding Red Hat OpenShift AI project-scoped resources: project-local workbench images, hardware profiles, and KServe model-serving runtime templates; dashboard visibility controlled by disableProjectScoped; resource naming and namespace rules; copying trusted global resources into a target project; and GitOps review of project-scoped resource manifests. Do NOT use for global RHOAI installation, generic OpenShift project RBAC, creating data science projects, custom workbench image build/import, full hardware profile schema design, full KServe serving runtime authoring, or live cluster changes without the OpenShift safety guard.

RHOAI Project-Scoped Resources

Use this skill to manage OpenShift AI resources that should be visible only in specific projects for the active baseline in docs/PLATFORM_BASELINE.md.

Source Grounding

Read references/source-capture.md before using product behavior details. Official Red Hat documentation is product authority. This skill adapts the official project-scoped resource workflow to this repo's GitOps review model.

Scope

This skill covers project-scoped copies of:

• workbench images
• hardware profiles
• model-serving runtimes for KServe, represented by Templates whose objects.kind is ServingRuntime

This skill does not cover full object schemas for those resources. Use the component owner skill and active CRD/schema verification for detailed fields:

• rhoai-project-workflows for creating data science projects and the user-facing project-scoped resource handoff workflow
• rhoai-workbenches-custom-images for workbench image content and dashboard discovery
• rhoai-hardware-profiles for hardware profile schema, lifecycle, and recommended accelerator tags
• rhoai-nvidia-gpu-accelerators for accelerator infrastructure intent
• rhoai-model-serving-platform for KServe ServingRuntime details and model-serving platform behavior

Resource Boundary

OpenShift AI users can access global resources across OpenShift AI projects. They can access project-scoped resources only in projects where they have permissions.

All resource names must be unique within a project. Project-scoped resources must set metadata.namespace to the target project.

Demo Policy

For this repo:

• Prefer project-scoped resources when a demo step needs different options for different teams, model-serving projects, or regulated audience scenarios.
• Keep global resources in the OpenShift AI application layer only when they should be visible across projects.
• Keep project-scoped resources in the same GitOps stage as the project that consumes them, or in a clearly ordered prerequisite stage.
• Set disableProjectScoped: false only when project-scoped resources should be enabled in the dashboard.
• Copy YAML only from trusted sources: official docs, existing verified resources, or reviewed GitOps manifests.
• Do not invent resource schemas. Verify ImageStream, HardwareProfile, Template, and embedded ServingRuntime fields before authoring.

Workflow

1. Confirm the active baseline in docs/PLATFORM_BASELINE.md.
2. Read references/official-doc-extraction.md.
3. Decide whether the target resource is a project-scoped:
   • workbench image
   • hardware profile
   • model-serving runtime template
4. Confirm project-scoped dashboard support is enabled with disableProjectScoped: false.
5. Copy YAML from a trusted source, update metadata.namespace, make metadata.name unique in the target project, and adjust only documented display-name fields.
6. Use examples/project-scoped-resource-patterns.md for review patterns.
7. Validate with references/validation-checklist.md.

References

• references/source-capture.md
• references/official-doc-extraction.md
• references/validation-checklist.md
• examples/project-scoped-resource-patterns.md