nat-path-checks
Use when fixing NeMo Agent Toolkit documentation path-check failures, especially failed `ci/scripts/path_checks.py` output, slash-delimited text mistaken for paths, relative path references, Markdown code escaping, and path-check allowlist decisions.
- License 许可
- Apache-2.0
- 最近更新
- 2026-08-31
- Stars
- ★ 0
- Forks
- 0
01 前言 Front Matter
- name
- nat-path-checks
- author
- NVIDIA Corporation and Affiliates
- license
- Apache-2.0
- description
- Use when fixing NeMo Agent Toolkit documentation path-check failures, especially failed `ci/scripts/path_checks.py` output, slash-delimited text mistaken for paths, relative path references, Markdown code escaping, and path-check allowlist decisions.
03 SKILL.md 正文
NeMo Agent Toolkit Path Checks
Use this skill when CI reports failed path checks from ci/scripts/path_checks.py, especially when copied CI output lists entries like:
Failed path checks:
- docs/example.md:40:10 -> data/test1
Workflow
- Open each reported file at the exact line before editing.
- Classify the reported token by meaning, not just by shape.
- Apply the smallest fix that preserves meaning.
- Re-run path checks and Markdown link checks on the changed files.
Fix Decision Table
| Reported token means | Preferred fix | Example |
|---|---|---|
| Prose shorthand, not a path | Rewrite in words. Do not hide it in code just to silence CI. | linux/amd64 -> linux or amd64; pass/fail -> pass or fail |
| Two field names or concepts | Name each item separately, usually as code spans if they are literal fields. | question/answer -> question and answer |
| Exact CLI, protocol, API, config, or method literal | Wrap the literal in inline code, or use a fenced code block for multi-line examples. | tools/list -> tools/list |
| Placeholder path in prose | Wrap the placeholder in inline code. | path/to/workflow.yml -> path/to/workflow.yml |
| Placeholder paths in a list or snippet | Use a fenced code block with an appropriate language or text. |
Put data/test1 and data/test2 inside a fenced block. |
| Real repo-relative path | Make the path correct relative to the current file, and link it when useful. | [README](../../README.md) |
| Intended generated output path | Use inline code or a fenced block unless the file should already exist in the repo. | ./output/results.json |
| Repeated false positive across many docs | Prefer a local wording or escaping fix. Add ALLOWLISTED_WORDS, IGNORED_PATHS, or ALLOWLISTED_FILE_PATH_PAIRS only for true checker limitations. |
Add an allowlist only after confirming the token should pass everywhere. |
Important Distinctions
- If a slash token is not meant to be copied exactly, rewrite it as prose.
- If a slash token is meant to be copied exactly, put it in Markdown code.
- If a slash token points to a real file in the repository, fix the target instead of escaping it.
- If
data/test1is an example input path, usedata/test1. If it is a real checked-in path, link to the real file or correct the path. - Do not replace a meaningful relative path with different wording unless the path was never meant to be literal.
Validation
Run these from the repository root after editing:
uv run ci/scripts/path_checks.sh
uv run pre-commit run markdown-link-check --files path/to/changed.md
If a copied CI log contains many failures, fix one file at a time and re-run path checks before applying the same pattern broadly.
04 同仓库更多 Skills
skill-installer
Install Codex skills into $CODEX_HOME/skills from a curated list or a GitHub repo path. Use when a user asks to list installable skills, install a curated skill, or install a skill from another repo (including private repos).
skill-creator
Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Codex's capabilities with specialized knowledge, workflows, or tool integrations.
openai-docs
Use for Codex models/pricing, scheduled tasks, skills, settings, setup, troubleshooting, customization, automations, and self-knowledge—including 'you,' 'your,' 'this app,' or 'this coding agent' when they refer to Codex—and for OpenAI APIs/products and ChatGPT Work. Also use for model choice/migration, prompting, SDKs, Responses, Realtime, agents, evals, and Chat/Work/Codex comparisons. Do not use for generic app/software tasks that merely mention Codex.
imagegen
Generate or edit raster images when the task benefits from AI-created bitmap visuals such as photos, illustrations, textures, sprites, mockups, or transparent-background cutouts. Use when Codex should create a brand-new image, transform an existing image, or derive visual variants from references, and the output should be a bitmap asset rather than repo-native code or vector. Do not use when the task is better handled by editing existing SVG/vector/code-native assets, extending an established icon or logo system, or building the visual directly in HTML/CSS/canvas.
skill-creator-cc
Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, update or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.
prd-writing-guide
Write complete, unambiguous PRDs that development teams can implement without guesswork. Includes requirement discovery framework, structured documentation methodology, completeness checklists, and common pitfall avoidance. Use when: writing new PRDs, reviewing PRD drafts, validating requirement completeness, preparing for engineering handoff. Triggers: 'write PRD', '写PRD', '产品需求文档', '需求文档', '需求规格', '需求评审', '完善需求', 'create requirements doc', 'product requirements', 'feature spec', 'requirements document'. Anti-triggers: 'technical design doc', 'architecture design', 'implementation plan', 'API design', '架构设计', '技术方案', '实现方案', '接口设计'.