菜单
Short chat reply via the ZeroGPU edge model (LFM2.5-1.2B-Instruct). Use when the user wants a quick, single-turn answer that does not need Claude-level reasoning, prior conversation context, or code generation. Optional system instructions via -i.
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。
基于 SOC 职业分类
Show how much you've saved by routing tasks to ZeroGPU instead of your frontier model — a rounded dollar estimate and the frontier-model tokens offloaded. Use when the user asks how much they've saved, their ZeroGPU savings, or to see the cost-savings summary.
Sign in to ZeroGPU and persist the API key. Use when the user asks to log in, authenticate, or set up ZeroGPU credentials for the first time.
Sign in to ZeroGPU and persist the API key. Use when the user asks to log in, authenticate, or set up ZeroGPU credentials for the first time.
Zero-shot classification against a caller-supplied list of candidate labels (deberta-v3-small). Use when the user wants to classify text into a custom set of labels they provide (e.g. "is this positive, negative, or neutral?", "tag this as bug, feature, or question").
Short chat reply via the ZeroGPU edge model (LFM2.5-1.2B-Instruct). Use when the user wants a quick, single-turn answer that does not need host-level reasoning, prior conversation context, or code generation. Optional system instructions via -i.
Chat with ZeroGPU's Thinking variant (LFM2.5-1.2B-Thinking), which returns visible reasoning alongside the answer. Use when the user wants the model's reasoning shown, or asks a short logic/math/word-problem question that benefits from step-by-step output.
| name | chat |
| description | Short chat reply via the ZeroGPU edge model (LFM2.5-1.2B-Instruct). Use when the user wants a quick, single-turn answer that does not need Claude-level reasoning, prior conversation context, or code generation. Optional system instructions via -i. |
| argument-hint | <text> [-i <instructions>] |
| allowed-tools | Bash(zerogpu chat *) |
Call the ZeroGPU chat model:
zerogpu chat $ARGUMENTS
Quoting (required, to survive shell parsing of arbitrary user text): format $ARGUMENTS with the prompt wrapped via heredoc command substitution, then any flags after. Pass the user's prompt verbatim inside the heredoc — do not paraphrase, do not escape:
"$(cat <<'ZGPU_T'
<the user's prompt here, verbatim, can span multiple lines and contain quotes/parens/$/etc.>
ZGPU_T
)" [-i "<system instructions>"]
Never inline the prompt as a plain "..." string — newlines, parens, single quotes, and $ in the user's text will break shell parsing. The heredoc form is safe for any input.
Savings note: only if the command output literally contains a line starting with 💰 ZeroGPU savings, append that exact line, unchanged, as the last line of your reply. If no such line is present, say nothing about savings and do not mention or suggest /zerogpu-router:cost-savings — this note is intentionally occasional, not shown every time.