用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/NVIDIA-Omniverse/omniverse-labs --skill usda-lexical-format命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
Use when setting up a fresh clone of this repo to run OVRTX Dev Variant Presenter (also called the "variant studio" in the walkthrough video and earlier releases) — no .venv yet, ovrtx/ovstage/ovstream not installed, ModuleNotFoundError on ovrtx/ovstage/ovstream, or "how do I install / get this running". RTX + uv; validated on Windows.
Use when asked to start, run, launch, restart, or bring up the OVRTX Dev Variant Presenter server, stream, or viewport in this repo — also called the "variant studio" in the walkthrough video and earlier releases — or after the stream / GPU / server was lost and needs recovering. RTX required; validated on Windows.
Build a browser-based, RTX-streamed app for exploring and rendering USD VARIANT PERMUTATIONS — a live ovstream/WebRTC viewport plus instant variant switching, cameras, a turntable rig, grid/batch stills, a multi-track timeline, projects, post, and remote-stage mirroring. Use when asked to build a "variant presenter / configurator" — or a "variant studio", the name used in the walkthrough video and earlier releases — a permutation render tool, or to add variant/timeline/batch features on top of an Omniverse realtime viewer.
基于 SOC 职业分类
正在显示 SKILL.md
| name | usda-lexical-format |
| description | Use this skill when implementing or verifying the USDA text-format substrate. |
| metadata | {"author":"NVIDIA"} |
Use this skill when implementing or verifying the USDA text-format substrate.
specification/file_formats/README.md section TextPinned tag / commit: v1.0.1
#usda layer header parsinginf, -inf, and nanThis skill owns contracts/capabilities/usda-lexical-format.json and provides
the usda_lexical_format capability consumed by USDA value and spec parsing.
It consumes the identifier_scanner capability from
usd-identifiers-and-names for identifier validity; it should not maintain an
independent UTF-8 or XID policy.
The lexer/parser substrate should preserve enough location information to emit line/column diagnostics, but exact diagnostic wording is controlled by goldens.
The substrate may be a lexer, parser-combinator layer, or direct recursive descent helpers. The contract is behavioral, not architectural.
Defer identifier UTF-8/XID scanning to usd-identifiers-and-names.
Defer path grammar after stripping angle brackets to usd-paths.
Defer value typing/coercion to usda-value-parser.
Do not map syntax to document-model fields here; that belongs to
usda-spec-parser.