chat
GitHub调用 ZeroGPU 的 gpt-oss-120b 模型进行聊天回复，适用于长文档、多步指令及复杂常识问题。支持通过 heredoc 安全传递提示词及可选系统指令，直接返回纯文本答案，并处理节省费用的特殊标记输出。
Trigger Scenarios
Install
npx skills add zerogpu/zerogpu-router --skill chat -g -y
SKILL.md
Frontmatter
{
"name": "chat",
"metadata": {
"openclaw": {
"install": [
{
"bins": [
"zerogpu"
],
"kind": "node",
"package": "zerogpu-cli"
}
],
"requires": {
"bins": [
"zerogpu"
]
}
}
},
"description": "Chat reply via ZeroGPU's default model, gpt-oss-120b (120B MoE, 131K context). Use when the user wants an answer from a ZeroGPU model rather than the host model, including longer documents, multi-step instructions, and harder general-knowledge questions. Optional system instructions via -i.",
"allowed-tools": "Bash(zerogpu chat_completions *)",
"argument-hint": "<text> [-i <instructions>]"
}
Sends your input to ZeroGPU's hosted API for inference. This is not local processing. Don't pass secrets, credentials, or regulated data you aren't cleared to share with a third party. See the plugin README's "Data & privacy" section.
Call the ZeroGPU chat model. Run this with the exec tool, pasting the user's prompt into the heredoc verbatim — no escaping or quoting required (the quoted heredoc handles every shell metacharacter, newline, quote, and paren safely):
zerogpu chat_completions -m gpt-oss-120b <<'ZGPU_END_OF_INPUT'
<the user's prompt, verbatim>
ZGPU_END_OF_INPUT
If the user supplied system instructions, add -i '<instructions>' after the model flag, writing any ' inside them as '\''.
Output is the assistant's answer as plain text. The model also produces a reasoning trace, which the Chat Completions response carries separately, so only the final answer is printed. Relay that answer as-is, without rewriting or expanding it.
For a faster, cheaper reply where quality matters less, use the chat-liquid skill (LFM2.5-1.2B-Instruct). For a visible reasoning trace, use chat-thinking. For multilingual prompts, use chat-qwen. When the input does not fit in this model's 131K context, use chat-deepseek for code and agentic work, on a 1M-token context, or chat-glm for the most capable option, on a 262K-token context — glm-5.2 costs roughly 7x this skill.
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 the cost-savings skill. This note is intentionally occasional, not shown every time.
Version History
-
4e1b070
Current 2026-09-22 08:07
v5.0.0：模型参数由117B更新至120B；v4.4.0：切换至 CLI 3.8.0 的模型无关端点命令，提升插件独立性。
-
87b63e9
2026-08-27 16:49
修正了 glm-5.2 相对于 chat 的成本倍数描述，从约 20x 更正为约 7x，以反映实际定价并优化路由决策。
-
cee9321
2026-08-02 21:15
将默认模型更新为gpt-oss-120b，原边缘模型功能移至chat-liquid；新增chat-glm和chat-deepseek技能以支持百万级上下文窗口。
- 7b9b503 2026-07-24 11:48
