← 返回 Skills 市场
274
总下载
0
收藏
2
当前安装
1
版本数
在 OpenClaw 中安装
/install user-permissions
功能描述
定义需要征求用户同意才能执行的操作规则。当任务涉及修改系统配置、安装/变更组件、执行可能影响系统运行的操作时，必须参考此skill确认是否需要先获得用户同意。特别适用于：系统管理任务、配置变更、skill安装/修改、定时任务管理、数据清理、网络配置、版本更新等场景。
安全使用建议
This skill is a permission/approval checklist and appears coherent and low-risk. Before installing: (1) Understand that it is instruction-only — it cannot enforce approvals by itself; the agent must be programmed to consult and follow it. (2) Test that your agent actually pauses and waits for explicit user confirmation for actions on the list (and verify what confirmation words are accepted). (3) Remember this skill does not block other skills from acting unless your agent integrates these checks into its execution flow. If you want enforced blocking, ensure the platform/agent enforces this policy centrally rather than relying solely on this instruction document.
功能分析
Type: OpenClaw Skill
Name: user-permissions
Version: 1.0.0
The 'user-permissions' skill bundle is a policy document (SKILL.md) designed to act as a safety guardrail for the OpenClaw agent. It explicitly instructs the agent to seek user consent before performing high-risk operations such as system configuration changes, software installation, cron job modification, or file deletion. There is no executable code, data exfiltration logic, or malicious intent; it serves to enhance system security by enforcing a 'consent-first' workflow.
能力评估
Purpose & Capability
The name and description match the SKILL.md content: the skill's sole purpose is to define when to pause and request user consent for side-effecting operations. It requests no environment variables, binaries, or installs that would be disproportionate.
Instruction Scope
The instructions are narrowly scoped to require user consent for a clear list of side-effecting operations (config changes, package installs, service control, file deletion, network changes, etc.). They reference expected system actions (e.g., apt, systemctl, rm -rf) only to enumerate what needs approval; they do not instruct the agent to read unrelated secrets or exfiltrate data. Note: the policy relies on the agent correctly recognizing when a task matches the checklist — the guidance is intentionally high-level and leaves detection/interpretation to the agent.
Install Mechanism
No install spec and no code files — instruction-only skill. This is the lowest-risk installation profile (nothing is written to disk).
Credentials
The skill declares no required environment variables, no credentials, and no config paths. There is no unexpected credential or secret access.
Persistence & Privilege
The skill is not 'always' enabled and does not request persistent or elevated platform privileges. It does not attempt to modify other skills or system configuration itself; it only instructs agents to pause and ask for consent.
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 user-permissions - 安装完成后，直接呼叫该 Skill 的名称或使用
/user-permissions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: user permission and approval rules
元数据
常见问题
User Permissions 是什么？
定义需要征求用户同意才能执行的操作规则。当任务涉及修改系统配置、安装/变更组件、执行可能影响系统运行的操作时，必须参考此skill确认是否需要先获得用户同意。特别适用于：系统管理任务、配置变更、skill安装/修改、定时任务管理、数据清理、网络配置、版本更新等场景。 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 274 次。
如何安装 User Permissions？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install user-permissions」即可一键安装，无需额外配置。
User Permissions 是免费的吗？
是的，User Permissions 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
User Permissions 支持哪些平台？
User Permissions 跨平台运行，可在任意部署了 OpenClaw / Claude Code 的环境中使用（cross-platform）。
谁开发了 User Permissions？
由 CCCU-IMU（@cccu-imu）开发并维护，当前版本 v1.0.0。
推荐 Skills