基于 SOC 职业分类
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/pytorch/helion --skill runpod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。
正在显示 SKILL.md
Investigate a Helion benchmark dashboard regression (helionlang.com/dashboard) — find the cause and classify it. Auto-activate when the user reports a perf drop/spike on a dashboard platform (e.g. b200 cute) around a given date.
Address CI failures and unresolved review comments on a Helion pull request. Auto-activate when the user mentions a URL like https://github.com/pytorch/helion/pull/<number>.
Context for working on Helion's TPU/Pallas backend. Auto-activate when the user mentions TPU or Pallas in the context of Helion development.
| name | runpod |
| description | Auto-activate when the user mentions runpod or scripts/runpod.py |
python scripts/runpod.py --start when you expect to need remote GPU runs. This avoids paying startup latency on every command.python scripts/runpod.py .... If a matching persistent pod is already running, the script reuses it automatically. If not, it falls back to one-shot start-run-stop behavior.python scripts/runpod.py HELION_BACKEND=cute pytest test.python scripts/runpod.py './lint.sh && pytest test'. Do not wrap it in bash -lc; the script handles shell execution itself.python scripts/runpod.py --cleanup so it does not keep billing while idle.