NVIDIA Elements is a design system and agent-oriented toolkit for building operational interfaces for AI/ML, robotics, and autonomous-vehicle software. Its CLI and MCP tools provide component APIs, design tokens, examples, imports, validation, and setup guidance across several web frameworks.
Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add NVIDIA/elements --skill audit-consolidationgit clone --depth 1 https://github.com/NVIDIA/elementsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/nvidia/elements/audit-consolidation)<a href="https://agentmods.dev/skills/nvidia/elements/audit-consolidation"><img src="https://agentmods.dev/badge/skills/nvidia/elements/audit-consolidation/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/nvidia/elements/audit-consolidation"><img src="https://agentmods.dev/badge/skills/nvidia/elements/audit-consolidation.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00126 | $0.01751 |
| Opus 5.5 | $0.00050 | $0.00700 |
| Sonnet 5.5 | $0.00025 | $0.00350 |
| Haiku 4.5 | $0.00013 | $0.00175 |
Grade A, and why
audit-consolidation scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 16d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 114 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Audit Consolidation
Find high-confidence changes that reduce the number of places a maintainer must understand or update for one concept. Favor cohesive ownership and simpler invariants over mechanically maximizing reuse.
Use audit-simplification instead when the primary question is whether state, indirection, lifecycle plumbing, synchronization, or another mechanism needs to exist at all. When both questions apply, run the simplification audit first; removing a mechanism often removes the duplication question with it.
Scope and Authorization
- Analyze the project or directory the user names. If they name no target, infer the smallest reasonable scope from the current request and repository state; ask only when materially different scopes are equally plausible.
- Treat the audit as read-only. Do not edit code merely because the audit finds an opportunity. Make a refactor only when the user explicitly asks for changes.
- Read the root
AGENTS.mdand any nearer agent instructions. Before analyzing a project, inspect itsDEVELOPMENT.md,package.json, architecture notes, public entrypoints, and directory structure when present. - Treat each project's
package.jsonexportsas its published boundary. Packages also export their/internalentrypoints, and sibling packages depend on them; for example, most other projects import@nvidia-elements/core/internal. Before calling anything unused or safe to move, search every project underprojects/, including starters, site docs,*.examples.tsfiles, andprojects/internals/patterns. - Preserve unrelated working-tree changes. Generated output, third-party code, dependency trees, build artifacts, snapshots, and coverage output are out of scope unless the user explicitly includes them.
Build an Evidence-Based Model
Start with a structural inventory, then follow behavior rather than relying on filenames or textual similarity alone.
- Map source directories, public and internal entrypoints, package boundaries, tests, and major dependency directions.
- Group code by the concept or invariant it owns: input normalization, validation, state transitions, resource lifecycle, data representation, rendering preparation, caching, serialization, error policy, or orchestration.
- Trace imports, exports, call sites, tests, and data flow for suspected overlaps. Search for synonymous names and operations with the same result, not only exact identifiers.
- Compare the implementations closely enough to identify their shared contract and intentional differences. When history would clarify why they diverged,
git log -S '<symbol>'finds the commits that added or removed a symbol, andgit log -L '<start>,<end>:<file>'follows one range of lines. - Check every proposed destination against package layering, public API, side-effect registration, dependency rules, and likely cycle creation.
projects/site/src/docs/api-design/packaging.mddescribes the entrypoint and side-effect conventions.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 16d ago First seen · 114 lines · 126 tokens per session scan A c114ac27ddcc
audit-consolidation is a skill published in the GitHub repository NVIDIA/elements (102 stars, last pushed today), licensed Apache-2.0. It adds 126 tokens to every session and 1,751 once invoked, about $0.0005 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-10-07.
Other skills, from other repositories
vue-application-structure
Establishes or reviews the directory layout, component conventions, composable design, Pinia store structure, and Vue Router configuration for a Vue 3 TypeScript application. Invoked when the user asks to structure a Vue app, set up the project layout, or review Vue architecture.
react-application-structure
Establishes or reviews the directory layout, feature boundaries, state design, routing approach, and data-fetching conventions for a React 18+ TypeScript application. Invoked when the user asks to structure a React app, set up a scalable architecture, or review React project organization.
next-application-structure
Establishes or reviews directory layout, server/client boundaries, routing, data-fetching strategy, and testing structure for Next.js 14+ App Router TypeScript applications. Invoked when the user asks to structure a Next app, set App Router conventions, or review architecture for React parity.
react-component-design
Designs or reviews React component APIs. Handles prop drilling decisions, composition patterns, controlled/uncontrolled contracts, and minimal public API surfaces for React 18+ TypeScript components. Invoked when creating new components, refactoring existing ones, or reviewing component contracts.
vue-component-design
Designs or reviews Vue 3 component APIs. Handles prop drilling decisions, applies compound component patterns with provide/inject, defines minimal public API surfaces with typed props/emits/slots, and enforces Vue 3.4+ composition conventions. Invoked when creating new components, refactoring existing ones, or…
compiler-commit
Use when you want to verify compiler changes and commit with the correct convention. Runs tests, lint, and format, then commits with the [compiler] or [rust-compiler] prefix.