CtrlK
BlogDocsLog inGet started
Tessl Logo

nim-setup

Configure NVIDIA NIM platform on OpenShift AI for optimized model inference. Use when: - "Set up NIM on my cluster" - "Configure NGC credentials for NIM" - "I want to deploy a NIM model but haven't set up the platform" - "Create the NIM Account CR" One-time prerequisite before deploying models with NVIDIA NIM runtime via /model-deploy. NOT for deploying models (use /model-deploy instead). NOT for vLLM or Caikit deployments (NIM-specific only).

66

Quality

81%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Passed

No findings from the security scan

SKILL.md
Quality
Evals
Security

Quality

Content

81%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A well-structured operational skill: the workflow is fully sequenced with mandatory confirmation gates and validation with recovery loops, and MCP tool usage is specified down to parameters and expected outputs. The main improvements are structural — move the inline Common Issues detail and the /model-deploy handoff notes into their reference files — plus showing how the dockerconfigjson payload is actually constructed.

Suggestions

Move the three detailed Common Issues subsections into references/common-issues.md, keeping only a one-line pointer in SKILL.md, since the body already links to that file.

Show the JSON structure of the .dockerconfigjson payload (e.g. {"auths":{"nvcr.io":{"username":"$oauthtoken","password":"<key>"}}}) before base64-encoding so the secret manifest is fully constructible.

Relocate the "NIM Deployment Handoff to /model-deploy" notes into a reference file or the /model-deploy skill, and trim the When-to-Use section that duplicates the frontmatter description.

DimensionReasoningScore

Conciseness

The body is dense operational content — MCP tool calls, parameter tables, manifests, and error branches — with essentially no explanation of concepts Claude already knows. Minor trimmable padding exists: the "When to Use This Skill" section largely restates the frontmatter description, and the three detailed Common Issues subsections duplicate a reference file the body already points to. Not 3 because the verbosity is minor and confined; not 5 because those redundant sections cost tokens without adding new information.

4 / 5

Actionability

Each step names the exact MCP tool, required parameters (apiVersion, kind, namespace, name), expected output, and error-handling branches, and manifests are near copy-paste ready ("manifest: full Secret manifest as JSON string - REQUIRED"). It falls short of 5 only because the .dockerconfigjson payload construction is never shown — the reader is told registry/username/password but not the JSON structure to base64-encode — and Step 7b's fallback tool lists parameters more loosely than the primary path.

4 / 5

Workflow Clarity

Steps 0–7 are explicitly sequenced, each creation step gates on "WAIT for explicit user confirmation" (a checklist-style HITL checkpoint), Step 1 validates operator health before proceeding, and Step 7 validates Account CR readiness and ServingRuntime creation with failure-driven feedback loops ("Regenerate NGC API key and re-run /nim-setup", "Wait 2-3 minutes and re-check"). This matches the top anchor: explicit validation steps, feedback loops for error recovery, and checkpoints for a fragile multi-step process.

5 / 5

Progressive Disclosure

The body is an overview with well-signaled, one-level-deep references (skill-conventions.md for shared prerequisites/HITL, supported-runtimes.md, common-issues.md, live-doc-lookup.md, examples/nim-setup.md — all present in the bundle). It does not reach 5 because content that belongs in those files is also inlined: three fully detailed Common Issues subsections sit alongside the pointer to common-issues.md, and the "/model-deploy handoff" notes are really guidance for another skill. Not 3: the split is mostly appropriate and every reference is clearly navigable.

4 / 5

Total

17

/

20

Passed

Description

82%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong description with explicit what/when structure, natural quoted triggers, and excellent boundary disambiguation. Its only weakness is that the capability statement stays at the level of one generic action ("Configure NVIDIA NIM platform") rather than naming the concrete artifacts the skill creates.

Suggestions

Rewrite the opening what-statement to enumerate the concrete actions, e.g. "Creates NGC image pull and API key secrets, the NIM Account custom resource, and verifies GPU/NFD operator health on OpenShift AI".

Add missing natural trigger variations such as "install NIM", "NIM prerequisites", or "NVIDIA NIM setup" to the Use-when list.

DimensionReasoningScore

Specificity

The what-statement is a single generic action ("Configure NVIDIA NIM platform on OpenShift AI for optimized model inference") rather than a list of concrete capabilities; concrete operations ("Configure NGC credentials", "Create the NIM Account CR") appear only inside the trigger list, matching the 'names domain and 1-2 concrete actions, but not comprehensive' anchor. It does not reach 4 because the descriptive body never enumerates the specific actions performed (create image pull secret, create API key secret, create Account CR, verify operators).

3 / 5

Completeness

Both what ("Configure NVIDIA NIM platform on OpenShift AI..." plus the one-time-prerequisite framing) and when (an explicit "Use when:" list of quoted trigger phrases) are clearly stated, with additional NOT-for boundaries. This matches the top anchor: explicitly answers both what AND when with concrete trigger phrases; below-anchor options lack either the explicit when-list or concrete triggers.

5 / 5

Trigger Term Quality

Natural user utterances are quoted directly ("Set up NIM on my cluster", "Configure NGC credentials for NIM", "I want to deploy a NIM model but haven't set up the platform") with good coverage of NIM/NGC/OpenShift phrasing. Not 5: common variations such as "install NIM", "NIM prerequisites", or "NVIDIA NIM" as a standalone phrase are missing.

4 / 5

Distinctiveness Conflict Risk

A clear niche (NIM on OpenShift AI) with explicit negative triggers — "NOT for deploying models (use /model-deploy instead)", "NOT for vLLM or Caikit deployments" — sharply separates it from the adjacent /model-deploy and vLLM/Caikit paths. Minimal conflict risk, matching the 'clear niche with distinct triggers' anchor.

5 / 5

Total

17

/

20

Passed

Validation

81%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 13 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

relative_links

Relative link issues: 1 deeper-than-1-level

Warning

referenced_paths_exist

Referenced path issues: 1 deeper-than-1-level

Warning

Total

13

/

16

Passed

Repository
RHEcosystemAppEng/agentic-plugins
Reviewed

Table of Contents

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.