nemotron-nano3

Reference desk for Nemotron 3 Nano / Llama-Nemotron Nano 3 — architecture, training data, recipes, evaluation, quantization, deployment. Use when the user asks facts about the model rather than building a pipeline.

Quality

79%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Fix and improve this skill with Tessl

tessl review fix ./skills/nemotron-nano3/SKILL.md

Quality

Content

71%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 retrieval skill body with a clear workflow, concrete routing table, and excellent progressive-disclosure file map. Its main weakness is redundancy across the Calibration Examples, Boundaries, and earlier guidance sections that inflates length without adding new guidance.

Suggestions

Trim or merge the Calibration Examples and Boundaries Do/Don't sections, which restate the tone and handoff guidance already given in Mission/Tone/Cross-Skill Handoff, to reduce token cost.

Add a one-line explicit validation checkpoint in the Retrieve step (e.g., 'confirm the cited file exists before answering') to strengthen workflow_clarity toward 5.

Collapse the Progressive Disclosure section's sequence into the Locate → Retrieve → Cite workflow to avoid restating the same navigation guidance twice.

Dimension	Reasoning	Score
Conciseness	Mostly efficient but carries redundant sections — the Calibration Examples and Boundaries Do/Don't restate the handoff and tone guidance from earlier sections — so it could be tightened, matching anchor 3; not 4 due to noticeable padding.	3 / 5
Actionability	Provides concrete, specific guidance: a question-type-to-file routing table, exact citation-format examples, and a handoff phrasing template covering common cases, fitting anchor 4; not 5 because there are no fully copy-paste artifacts, which is appropriate for an instruction-only retrieval skill.	4 / 5
Workflow Clarity	A clear 'Locate → Retrieve → Cite' sequence with numbered ordered substeps and an explicit Source Priority ordering; checkpoints are implicit rather than explicit, matching anchor 4. Not a destructive/batch operation, so no validation cap applies.	4 / 5
Progressive Disclosure	Serves as a clear overview with a closing file-tree map and well-signaled one-level-deep references to paper/, recipes/, model-card.md, and context/, making navigation easy per anchor 5.	5 / 5
	Total	16 / 20 Passed

Description

87%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 tight, third-person description that names the model, enumerates concrete coverage areas, and gives an explicit trigger with a carve-out against a sibling skill. Minor specificity gap is that capabilities are topics rather than verb-phrased actions.

Dimension	Reasoning	Score
Specificity	Lists six concrete capability areas (architecture, training data, recipes, evaluation, quantization, deployment), matching anchor 4 for several specific actions with minor gaps; not 5 because these are topical areas rather than verb-phrased actions.	4 / 5
Completeness	Explicitly answers both what ('Reference desk for…' with enumerated areas) and when ('Use when the user asks facts about the model rather than building a pipeline'), matching anchor 5.	5 / 5
Trigger Term Quality	Includes a natural trigger phrase ('Use when the user asks facts about the model') and model-specific terms, but is missing some common synonyms/extensions, fitting anchor 4.	4 / 5
Distinctiveness Conflict Risk	Occupies a clear model-specific niche with an explicit contrast carve-out ('rather than building a pipeline'), giving minimal conflict risk per anchor 5.	5 / 5
	Total	18 / 20 Passed

Validation

100%

Warnings & errors only

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

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: NVIDIA-NeMo/Nemotron
Path: skills/nemotron-nano3/SKILL.md
Commit: 1a4c3fa

Reviewed: 4 days ago

Table of Contents

Discovery Implementation Validation

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.