elements AGENTS.md

A repository instruction guide for coding agents working on NVIDIA’s Elements project, a monorepo that contains several related packages in one codebase. It explains which instructions take priority, how the project is organised, and how to work with it.

In plain words
What is it for?
Following repository policies, understanding the Elements codebase, checking environment requirements, and running common development commands.
Why use it?
It reduces mistakes caused by conflicting instructions or missing setup knowledge. Agents can follow the project’s required rules and commands in the right order.

Instructions file for CodexOpenCode

Install

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.

agentmods
npx agentmods add instructions/nvidia/elements/agents-md
Clone the repo
git clone --depth 1 https://github.com/NVIDIA/elements

Made for: Codex, OpenCode.

Per session 3,066 This file is loaded in full into every session.
When invoked 3,066 The same file — it is already loaded in full.
Security scan C 2 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5 $0.03066 $0.03066
Opus 5 $0.01533 $0.01533
Sonnet 5 $0.00613 $0.00613
Haiku 4.5 $0.00307 $0.00307

Measured 2d ago against content hash 0aa4db324b02, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade C, and why

elements AGENTS.md scanned grade C with 2 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 2d 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.

Downloads and executes remote codehighSupply chain

curl | sh runs whatever the server returns today, which is not necessarily what it returned when this was reviewed.

curl https://mise.run | sh

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

curl https://mise.run | sh
AGENTS.md · 285 lines

How it starts

The opening of the file, as written. The whole thing — 285 lines — stays where its author put it; the contents beside it link to each section on GitHub.

AGENTS.md

This file provides guidance to AI Agents when working with code in this repository.

Instruction Precedence for Agents

When instructions conflict, follow this order (highest to lowest):

  1. Runtime or platform constraints from the agent host
  2. Repository agent rules in AGENTS.md
  3. Task-specific skill instructions in .claude/skills/*/SKILL.md
  4. Documents explicitly required by a selected skill (for example, docs in projects/site/src/docs/internal/guidelines/)
  5. General project documentation such as README.md

Notes:

  • README.md is primarily for orientation and setup; it does not override agent policy.
  • If a skill says you MUST review a referenced guideline, treat that guideline as required for that task.
  • If two same-level sources conflict, prefer the more specific and recently maintained source, and note the assumption in your response.

Repository Overview

Elements is a design language for AI/ML factories built as a monorepo containing framework agnostic Web Components (Lit), themes, styles, testing utilities, and starter templates. The repository uses pnpm workspaces with Wireit for build orchestration and Semantic Release for automated versioning/publishing.

Environment Requirements

  • mise: Toolchain manager for local development and CI (mise.toml)
  • Node.js: 26.7.0 (managed via mise and validated against .nvmrc and package.json engines)
  • pnpm: 11.23.0 (managed via mise and validated against package.json)
  • Go: 1.26.4 (managed via mise for the Go starter)
  • Git LFS: Required for visual test screenshots and videos (.gitattributes defines tracked files)
  • Playwright: Browser-based testing uses Chromium (installed via prepare script)
  • Vale: Prose linter for documentation and JSDoc (managed via mise)

Common Commands

Repository Setup

# Install required tools and run CI
curl https://mise.run | sh
# If mise is not on PATH yet, use ~/.local/bin/mise for the setup commands.
mise trust
mise run setup

Read the full file on GitHub · 285 lines

Changes

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.

  1. 2d ago First seen · 285 lines · 3,066 tokens per session scan C 0aa4db324b02

Subscribe to this mod's changes

elements AGENTS.md is an instructions file published in the GitHub repository NVIDIA/elements (83 stars, last pushed 2d ago), licensed Apache-2.0. It adds 3,066 tokens to every session, about $0.0153 per session on Opus 5. A static security scan graded it C with 2 findings (downloads and executes remote code, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.