复制安装命令
用 Codex 或 Claude 安装复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它先审查 Skill 页面再帮你安装。
复制前请先查看来源、License 和安全提示。
SkillAtlasSkill 详情
rtx-remix-modding
Official, NVIDIA-verified Agent Skills for Claude Code, Codex, and other coding agents.
审核状态：已审核Quality 80Security 70
项目 README
来源文件：README.md
NVIDIA Agent Skills
Official, NVIDIA-verified Agent Skills for Claude Code, Codex, and other coding agents.
📖 Docs: docs.nvidia.com/skills · 📺 Livestream: From Vulnerable to Verified · 📝 Blog: NVIDIA Verified Agent Skills: Capability Governance for AI Agents
Skills are portable instruction sets that teach AI agents how to use NVIDIA software optimally: Physical AI and robotics workflows, simulation, CUDA-X libraries, RAG and AI Blueprints, and platform tools. This repository is a catalog: skills are maintained in their respective product repos, and mirrored here daily via an automated sync pipeline. Skills are being added continuously, so check back for updates. We are building this infrastructure in the open, and contributions are welcome. See the Roadmap for what is planned next.
Quickstart
Install NVIDIA skills with the default skills CLI flow:
npx skills add nvidia/skills
The CLI runs through npx and prompts you to choose a skill and install destination. You do not need to clone this repo or copy skill folders by hand.
Requires a current
skillsCLI (v1.5.16 or newer). Installing vianpx skills@latest add nvidia/skillsalways uses the latest. On older CLIs (v1.5.15 and earlier), skills may install but not appear in Claude Code — see Troubleshooting.
The skill is available the next time your agent loads skills and encounters a relevant task. For example, ask your agent to "solve a linear programming problem with cuOpt" and the skill guides it through the cuOpt Python API. In Claude Code, run /reload-skills to load newly installed skills in your current session.
Install One Skill Without Prompts
Use this when you already know the skill name and want to skip prompts.
npx skills add nvidia/skills --skill cuopt-numerical-optimization-api --yes
Replace cuopt-numerical-optimization-api with any skill name from the Skill Catalog.
Install for a Specific Agent
Use --agent to target a specific AI coding agent. Initially, we'll support common client targets, expanding the list over time. For the full list of clients supported by the spec, see the skills CLI Supported Agents table.
Claude Code
npx skills add nvidia/skills --skill cuopt-numerical-optimization-api --agent claude-code
Codex
npx skills add nvidia/skills --skill cuopt-numerical-optimization-api --agent codex
Snowflake CoCo
npx skills add nvidia/skills --skill cuopt-numerical-optimization-api --agent cortex
Cursor
npx skills add nvidia/skills --skill cuopt-numerical-optimization-api --agent cursor
Kiro
npx skills add nvidia/skills --skill cuopt-numerical-optimization-api --agent kiro-cli
Use --agent more than once to install the same skill into multiple agents.
npx skills add nvidia/skills \
--skill cuopt-numerical-optimization-api \
--agent claude-code \
--agent codex \
--agent cursor \
--agent kiro-cli
Keep Skills Up to Date
New skills land continuously, and existing ones are revised, renamed, or consolidated as the catalog evolves. Refresh what you have installed with:
npx skills update
Run it interactively and the CLI also flags skills that were removed or merged upstream (for example, when several skills are consolidated into one) and offers to remove the stale local copies. Use npx skills list to see what is installed and npx skills check to preview what is out of date first.
Browse the Catalog
Use this when you want to see available NVIDIA skills before installing anything.
npx skills add nvidia/skills --list
For non-interactive installs, global installs, agent-specific installs, updates, removals, and fallback manual copying, see Advanced installation.
Skill Catalog
Getting Help & Contributing
Where to file an issue depends on what's broken:
- Skill content issues (a specific skill has a bug, missing functionality, or incorrect content) — file in the source repo for that product, using the per-product table below.
- Catalog issues (catalog README errors, sync workflow problems, distribution channels, signing/verification flow, docs in this repo) — file here using the catalog issue templates: Bug Report, Feature Request, or Documentation Request or Correction.
- Questions or general discussion — use Discussions. The issue tracker is reserved for bug reports, feature proposals with a design, and documentation issues.
- Security vulnerabilities — follow the disclosure process in SECURITY.md; do not open a public issue.
Per-product source repo links:
For issues with this catalog repo itself (README, structure, listing a new product): open an issue here.
Verifying Skills
Every published skill ships with a detached OMS signature (skill.oms.sig). The sync pipeline drops any skill missing the required artifacts before publishing, so every skill in the catalog carries:
SKILL.md— the skill instructions consumed by the agentskill-card.md— skill identity and governance cardskill.oms.sig— detached OMS signature (verifiable againstnv-agent-root-cert.pem)- A Tier-3 evaluation dataset — accepted at
evals/evals.json,evals/*.json,eval/*.json, orbenchmark/evals.json BENCHMARK.md— generated benchmark report capturing verifiable uplift data
Verify a skill against the NVIDIA trust anchor nv-agent-root-cert.pem:
pip install model-signing
model_signing verify certificate SKILL_DIR \
--signature SKILL_DIR/skill.oms.sig \
--certificate_chain nv-agent-root-cert.pem \
--ignore_unsigned_files
A successful verification confirms that the skill contents have not been modified since signing by NVIDIA.
See Verify Signed Agent Skills for signature layout, the trust pipeline, and policy options.
Roadmap
- ✅ Public skills catalog with NVIDIA-verified skills across multiple products
- ✅ Automated sync pipeline with skills mirrored from product repos daily
- ✅ Security scanning for all published skills covering instruction safety and supply-chain integrity
- ✅ Skills signing so every published skill carries a verifiable NVIDIA signature
- ✅ Skills universal evaluation criteria and task-specific criteria
- ✅ Skill Card with machine-readable metadata for identity, provenance, quality, and behavioral boundaries
- ✅ Sync-time compliance gates — signature drift detection and missing-artifact enforcement
- ✅ Syndication to external marketplaces — Skills.sh, Codex plugin, Claude Code plugin, ClawHub, Hermes Hub
- 🔲 Syndication to additional MCP hubs and partner channels
Repository Structure
NVIDIA/skills/
├── skills/ # NVIDIA-verified skills (count grows continuously),
│ │ synced from upstream product repos
│ ├── README.md # Browser-facing install guidance
│ ├── <product-prefix>-*/ # Flat layout — one dir per skill, product-prefixed
│ │ # e.g. aiq-*, cuopt-*, cupynumeric-*,
│ │ # dali-*, deepstream-*, dicom-*, digital-health-*,
│ │ # dynamo-*, earth2studio-*, holoscan-*, hsb-*,
│ │ # jetson-*, launch-nemo-rl, mcore-*,
│ │ # nemo-automodel-*, nemo-data-designer-plugin,
│ │ # nemo-evaluator-plugin, nemo-mbridge-* (20 skills),
│ │ # nemo-retriever, nemo-rl-* (4 skills),
│ │ # nemoclaw-user-guide, nemotron-*, nemotron-speech,
│ │ # nv-* (medical AI), physicsnemo-*, rag-*,
│ │ # skill-card-generator, tao-*, tilegym-*,
│ │ # vss-* (15 skills), accelerated-computing-cudf,
│ │ # cudaq-guide, portfolio-optimization
│ ├── omniverse-*/ # Physical AI — manually staged (see manual-components.yml)
│ └── physical-ai-*/ # Physical AI — manually staged
├── components.d/ # Product registry — one file per component, teams onboard here
│ ├── README.md # Schema and onboarding instructions
│ └── <product>.yml # one file per registered product
├── plugins/ # Packaged plugin distributions
│ └── nvidia-skills/ # Curated NVIDIA skills bundle (Claude Code, Codex)
├── plugins.d/ # Plugin build registry — config for `build-plugins.py`
│ ├── README.md
│ ├── _defaults.yml
│ └── nvidia-skills.yml
├── .claude-plugin/ # Claude Code marketplace metadata
│ └── marketplace.json
├── .agents/plugins/ # Agent marketplace metadata (other clients)
│ └── marketplace.json
├── docs/ # Long-form documentation (published via Fern)
│ ├── README.md # How to build the docs locally
│ ├── index.mdx
│ ├── advanced-install.mdx
│ ├── agent-skill-trust-pipeline.mdx
│ ├── release-checklist.mdx
│ ├── scanning-agent-skills.mdx
│ ├── signing-agent-skills.mdx
│ └── skill-cards.mdx
├── fern/ # Fern docs site configuration
├── .github/
│ ├── workflows/ # Sync pipeline, plugin validation, DCO check, author verify
│ └── scripts/ # regenerate-readme.sh, build-plugins.py,
│ # manual-components.yml (temp Physical AI catalog
│ # exception, removed after Computex 2026),
│ # marketplace/metadata.json (skill metadata sidecar)
├── nv-agent-root-cert.pem # Trust anchor for OMS signature verification
├── skills.sh.json # Skills.sh marketplace grouping config
├── CHANGELOG.md
├── CONTRIBUTING.md # Contribution guidelines
├── SECURITY.md # Security reporting policy
├── CODE_OF_CONDUCT.md # Community code of conduct
├── LICENSE-APACHE # Apache 2.0 (source code)
└── LICENSE-CC-BY-4.0 # CC BY 4.0 (documentation/skills)
Skills are maintained in their respective product repos (see the Source column in the Skill Catalog) and synced to this repo daily. Products only appear under skills/ after the sync pipeline confirms each skill carries:
skill.oms.sig— detached OMS-format signature (verifiable againstnv-agent-root-cert.pem)skill-card.md— skill identity and governance card- A Tier-3 evaluation dataset — accepted at
evals/evals.json,evals/*.json,eval/*.json, orbenchmark/evals.json
When evaluation runs produce a BENCHMARK.md, it ships alongside the skill so consumers can see verifiable benchmark uplift data.
Standards & Compatibility
This repository adheres to the Agent Skills specification:
- Skills are portable directories with a
SKILL.mdfile at their root. - Metadata uses YAML frontmatter with required
nameanddescriptionfields. - Skills follow a progressive disclosure model — lightweight metadata loads at startup, full instructions load on activation.
- Validate your skill using the
skills-refreference library.
License
Copyright (c) 2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved.
This code is dual-licensed with documentation/skills under the CC-BY-4.0 AND source code under Apache-2.0 license terms. The full license texts can be found in LICENSE-APACHE and LICENSE-CC-BY-4.0 respectively.
Agent / MCP / Skill 创作
高风险
- 来源需自行核对维护者身份。
- 未检测到明显脚本安装指令。
- 未检测到明显外部权限要求。
- 存在潜在风险命令，请谨慎安装。
- 扫描发现：3 条。
Codex — Git Clone 安装
- 安装前请先查看来源仓库和风险报告。
- 克隆仓库：git clone https://github.com/NVIDIA/skills.git
- 将 "skills/rtx-remix-modding" 文件夹复制到 Codex 的 skills 目录中。
- 重启 Codex 让新的 skill 生效。
Codex — 手动复制安装
- 安装前请先查看来源仓库和风险报告。
- 从源仓库下载 SKILL.md 及相关文件。
- 在 Codex 的 skills 目录中创建新文件夹。
- 将所有 skill 文件复制到新文件夹中。
- 重启 Codex 让新的 skill 生效。
Claude Code — Git Clone 安装
- 安装前请先查看来源仓库和风险报告。
- 克隆仓库：git clone https://github.com/NVIDIA/skills.git
- 将 "skills/rtx-remix-modding" 文件夹复制到 Claude Code 的 skills 目录中。
- 重启 Claude Code 让新的 skill 生效。
Claude Code — 手动复制安装
- 安装前请先查看来源仓库和风险报告。
- 从源仓库下载 SKILL.md 及相关文件。
- 在 Claude Code 的 skills 目录中创建新文件夹。
- 将所有 skill 文件复制到新文件夹中。
- 重启 Claude Code 让新的 skill 生效。
Cursor — Git Clone 安装
- 安装前请先查看来源仓库和风险报告。
- 克隆仓库：git clone https://github.com/NVIDIA/skills.git
- 将 "skills/rtx-remix-modding" 文件夹复制到 Cursor 的 skills 目录中。
- 重启 Cursor 让新的 skill 生效。
Cursor — 手动复制安装
- 安装前请先查看来源仓库和风险报告。
- 从源仓库下载 SKILL.md 及相关文件。
- 在 Cursor 的 skills 目录中创建新文件夹。
- 将所有 skill 文件复制到新文件夹中。
- 重启 Cursor 让新的 skill 生效。
GitHub Copilot — Git Clone 安装
- 安装前请先查看来源仓库和风险报告。
- 克隆仓库：git clone https://github.com/NVIDIA/skills.git
- 将 "skills/rtx-remix-modding" 文件夹复制到 GitHub Copilot 的 skills 目录中。
- 重启 GitHub Copilot 让新的 skill 生效。
GitHub Copilot — 手动复制安装
- 安装前请先查看来源仓库和风险报告。
- 从源仓库下载 SKILL.md 及相关文件。
- 在 GitHub Copilot 的 skills 目录中创建新文件夹。
- 将所有 skill 文件复制到新文件夹中。
- 重启 GitHub Copilot 让新的 skill 生效。
Windsurf — Git Clone 安装
- 安装前请先查看来源仓库和风险报告。
- 克隆仓库：git clone https://github.com/NVIDIA/skills.git
- 将 "skills/rtx-remix-modding" 文件夹复制到 Windsurf 的 skills 目录中。
- 重启 Windsurf 让新的 skill 生效。
Windsurf — 手动复制安装
- 安装前请先查看来源仓库和风险报告。
- 从源仓库下载 SKILL.md 及相关文件。
- 在 Windsurf 的 skills 目录中创建新文件夹。
- 将所有 skill 文件复制到新文件夹中。
- 重启 Windsurf 让新的 skill 生效。
查看 SKILL.md 原文
name: rtx-remix-modding
description: Mod or remaster a game with RTX Remix - open and edit projects, swap textures and models. Connect to the Remix Toolkit App via MCP. Not for non-Remix game interaction.
metadata:
author: scfitzpatric <scfitzpatric@nvidia.com>
tags:
- rtx-remix
- game-modding
- usd
- mcp
- graphicsRTX Remix Modding
When to Use This Skill
Use this when the user is modding or remastering a classic DirectX 8/9 game with NVIDIA RTX Remix and a Remix MCP server is available. It covers opening and closing projects, working with the project's USD layers, swapping the model and texture assets bound to captured prims, and reading the stage. It does not cover installing Remix, capturing a game, or authoring new geometry.
Connecting to Remix
Requires RTX Remix 1.6.0+ or 1.6.0-dev; ask the user to upgrade if older.
Missing remix_* tools? Ask the user to start Toolkit, windowed or via
lightspeed.app.trex.stagecraft.headless.bat, and connect to http://127.0.0.1:18014/mcp/.
Keep the trailing slash. Both modes fall back within 18014–18019; all busy means startup fails.
Windows discovery record, published by Remix:
%LOCALAPPDATA%\NVIDIA\RTX Remix\mcp.json.
Read mcp_endpoint for the latest instance. Contains addresses, ports, version, and PID;
removed on clean shutdown.
No record? Use endpoint= from the Toolkit log's MCP_PORT_FALLBACK warning.
Wait for matching SERVICE_READY service=mcp. Client URLs need manual updating.
Instructions
Before using the MCP tools to replace models or add an asset reference, read
references/tool-mode.md and follow the relevant recipe.
Your tool list already carries every tool and its schema. What it does not carry is the order:
- Open or confirm the project.
- Set the edit target before writing. A write lands in whichever layer is current, and the wrong layer is the failure that looks like success.
- Find the prims —
selection=trueacts on the viewport selection. - Write. A file has to be ingested before it can be referenced.
remix_save_layer, then read back what you changed.
Troubleshooting
Before planning, check for remix_* tools; if absent, say the server is unavailable and relay
the setup instructions in "Connecting to Remix".
Never substitute a workaround for the missing tools. Editing .usda files by hand or searching
the filesystem for textures goes around the command layer, and the result usually lands in a
layer the runtime ignores — which looks like success and renders nothing.
When a call fails, read the response before trying again. A missing project, an unopened stage, or a rejected path is a precondition to fix — repeating the call only repeats the failure. Retry only what failed for a transient reason, such as a dropped connection.
Rules
Scope caveat you must always disclose: edits apply per captured asset, not per instance, so every instance sharing that mesh changes too. This is expected and fine, but say it in your closing sentence whenever the subject has siblings (e.g. "Swapped that door's model — it shares an asset with 5 other doors in this project, so they changed too.").
Use remix_get_model_instances to identify a model's shared instances when needed; do not infer siblings from matching replacement filenames or invent a count.
There is no undo tool in this release. Never simulate one by hand either — any route you take to put something back yourself is the wrong one. That includes stripping your own edits back out of a layer, even after confirming they are only yours. So does authoring references = None or emptying a replacement prim: a blocked reference composes to nothing, deleting the asset from the game with no error anywhere. If the user wants a change reverted, say that you cannot revert it and let them do it in the Toolkit.
What you can and cannot do, by thing. Nothing outside this exists — no tool and no workaround. Say so, and offer the nearest thing you can do:
- Layers — the only thing you can create or delete.
- Projects — open and close an existing one. You cannot create one.
- Prims — read them and set the selection. You cannot create, delete, rename, duplicate or move one. That covers geometry of every kind: no meshes, no new objects, by any route.
- Asset references — add one to a prim, or replace one already there. There is no removal.
- Textures — read what a material binds, and replace it.
Nothing saves itself in this release. A write lands in memory and the running game keeps rendering the old asset until the layer is on disk, so a turn that ends on a successful edit has changed nothing the user can see. Call remix_save_layer once the edit is in, then read back to confirm it landed. This is the same silent success as above, and the easier one to report as a win.
Ask instead of guessing scope. "Swap all the broken ones", "fix everything in this layer" — ask how many, and which, before acting.
Stop repeating a failing approach. Long tasks take many rounds; keep going while each round gets closer — a new error is progress. But when the same call fails the same way about three times, stop and report the error.
A call that succeeded and told you nothing useful is the same dead end, and it is the easier one to miss because nothing looks broken. NEVER re-run a query you have not invalidated. If nothing has happened since it last ran, it will return exactly what it returned, and running it a third time spends your turn budget on an answer you already hold. This covers searching the stage as much as anything else — a name that was not there is still not there. When a search comes back empty, that emptiness IS your answer: report what you looked for, report what you DID find, and ask the user to name the thing another way. Ending your turn with a question is a correct outcome, not a failure, and it beats spending every remaining call re-asking something you have already answered.
A tool saying the thing is not there is the WHOLE answer, not the first of two opinions. Those tools read the same stage you would walk yourself, so asking a second one returns the same nothing — and hunting through prim children for a texture the material does not carry, or a name the capture does not hold, is the loop, not the way out of it. Plenty of what a game renders is not in a capture at all: sky, fog, water, shadows, the HUD. Say what you found and that you cannot reach it. One thing DOES invalidate a query, and re-running is right for it: you changed the scene — read back after a write to confirm it landed.
If no tool can do what the user asked, say so and stop. "I can't do that with the tools I have — here is what I can do instead" is a correct, useful answer. Improvising around a missing capability is how real damage happens; a plain refusal costs the user one turn and nothing else.
When a tool refuses, RELAY WHAT IT TOLD YOU. These refusals are written for the user, not for you: they name the specific thing that blocked the call and, where one exists, the remedy. Pass both on verbatim in your own sentence. Two failures to avoid — dropping the remedy, so the user hears "impossible" when the tool said "do X and this works"; and explaining the CAUSE yourself when the tool did not give you one. If you find yourself writing "likely" or "probably" about why something is blocked, you are guessing at the user's setup. Say what the tool reported and what would unblock it, and stop there.
When you have finished the user's request, end your turn with one short, plain-language sentence telling the user what you did or answering their question — for an action, confirm it (e.g. "Opened your most recent project.").
Short never means contentless. "Done." and "OK" are not acceptable answers — they tell the user nothing and hide whether the thing they asked for actually happened. Name what changed, and if any rule above told you to state a caveat or a refusal, that caveat is the sentence. Say it even when it is the only thing you have to report.
评论 (0)
暂无评论，成为第一个评论者吧！