← 返回 Skills 市场
445
总下载
0
收藏
2
当前安装
1
版本数
在 OpenClaw 中安装
/install qwen
功能描述
Build and route Qwen chat, coding, reasoning, and vision workflows across hosted and self-hosted endpoints with safer debugging.
安全使用建议
This skill appears coherent with its stated purpose. Before installing, confirm you are comfortable with the agent creating and storing short routing/memory files under ~/qwen/ (it says it will ask first). Provide DASHSCOPE_API_KEY only if you plan to use the hosted DashScope/Model Studio endpoints — it isn't necessary for local-only setups. Verify the listed external endpoints (dashscope*.aliyuncs.com) match the service you expect to use for hosted models, and keep your API key in environment variables as advised.
功能分析
Type: OpenClaw Skill
Name: qwen
Version: 1.0.0
The Qwen skill bundle is a well-documented tool for routing and managing Qwen AI model workflows across hosted (Alibaba DashScope) and local (Ollama, vLLM) environments. It uses standard system utilities (curl, jq) for API interactions and follows security best practices by instructing the agent to use environment variables for secrets and seek user approval before persisting state to ~/qwen/. No evidence of malicious intent, data exfiltration, or unauthorized execution was found.
能力评估
Purpose & Capability
The name/description (Qwen routing across hosted and self-hosted endpoints) matches the instructions and files: examples show hosted DashScope endpoints and local OpenAI-compatible servers. Required binaries (curl, jq) are used in health-check and minimal-request snippets. The declared config path (~/qwen/) is used for routing/memory files. Overall the requested resources are appropriate for the stated purpose.
Instruction Scope
SKILL.md and the included docs confine actions to model discovery, chat completions, local health checks, and optional local file writes under ~/qwen/. The skill explicitly lists the dashscope.* endpoints as the only external endpoints and instructs to ask the user before creating persistent notes. There are no instructions to read unrelated system files or to send data to undeclared third parties.
Install Mechanism
This is an instruction-only skill with no install spec or downloads; that is the lowest-risk install mechanism. No archives, remote installers, or third-party package pulls are present.
Credentials
Only one environment variable is requested: DASHSCOPE_API_KEY, which is appropriate for the documented hosted DashScope / Model Studio endpoints. Minor incoherence: the registry metadata lists DASHSCOPE_API_KEY as a required env var even though the skill supports self-hosted flows (Ollama, vLLM, etc.) that do not need that key. It would be more precise to mark the key as required only for hosted workflows or make it optional when the user intends local-only usage.
Persistence & Privilege
The skill does not request elevated privileges or permanent/always-on inclusion. It declares a config path (~/qwen/) and instructs to create files there only after asking the user. always:false and normal model invocation are used; nothing in the docs indicates modification of other skills or system-wide settings.
如何使用
- 确保已安装 OpenClaw（本地或 Docker 部署）
- 在对话框中输入安装命令：
/install qwen - 安装完成后，直接呼叫该 Skill 的名称或使用
/qwen触发 - 根据 Skill 的参数说明提供必要输入，即可获得结构化输出
版本历史
v1.0.0
Initial release with hosted and self-hosted Qwen routing, API patterns, tool-calling guidance, and troubleshooting playbooks.
元数据
常见问题
Qwen 是什么？
Build and route Qwen chat, coding, reasoning, and vision workflows across hosted and self-hosted endpoints with safer debugging. 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 445 次。
如何安装 Qwen？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install qwen」即可一键安装，无需额外配置。
Qwen 是免费的吗？
是的，Qwen 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
Qwen 支持哪些平台？
Qwen 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（linux, darwin, win32）。
谁开发了 Qwen？
由 Iván（@ivangdavila）开发并维护，当前版本 v1.0.0。
推荐 Skills