by 安全扫描
OpenClaw
安全
high confidenceThe skill's files and runtime instructions match its stated purpose (a preflight/runtime lifecycle guard); it requires no credentials, no installs, and its behavior is advisory and self-contained.
评估建议
This skill appears internally consistent and low-risk: it contains guard policies and checklists and does not request credentials, install code, or reach out to external URLs. Before enabling: (1) confirm you trust the skill author (owner ID is present but no homepage or repo was provided); (2) review the included files (they are present in the package) to satisfy yourself the audit rules are acceptable; (3) be aware the guard will read all files inside the skill package during preflight and wil...详细分析 ▾
✓ 用途与能力
Name/description (lifecycle guard, route to preflight/runtime) align with the included artifacts: preflight and runtime guard documents and stage checklists. There are no unrelated requirements (no env vars, no binaries, no installers) that would contradict the stated purpose.
✓ 指令范围
All runtime instructions are guard/advisory rules (triage before tool calls, stage guards, preflight artifact review). The preflight explicitly limits file reads to the candidate skill's own directory and prohibits traversing system paths. The instructions do not ask the agent to read or exfiltrate data outside the skill package, nor to contact external endpoints.
✓ 安装机制
This is an instruction-only skill with no install spec and no code to execute. Nothing is written to disk by an installer; risk from supply-chain installation steps is minimal.
✓ 凭证需求
The skill declares no required environment variables, credentials, or config paths. The preflight guard documents how to treat references to external paths as findings but does not read them—this is proportionate to auditing a skill package.
ℹ 持久化与权限
always:false and no install means the skill does not demand permanent presence. However, the skill is allowed to be invoked autonomously (disable-model-invocation:false). Because it's a guard, autonomous invocation could affect agent behavior (block actions, require confirmations) across sessions; this is expected for a policy guard but is operationally impactful rather than a security incoherence.
安全有层次，运行前请审查代码。
运行时依赖
无特殊依赖
版本
latestv1.2.12026/3/30
- Clarified the distinction between preflight and runtime guard modes. - Outlined decision rules for ambiguous cases and user overrides. - Added constraints for language support, advisory nature, on-demand loading, and risk escalation. - Improved documentation in SKILL.md for easier understanding of guard logic.
● 无害
安装命令 点击复制
官方npx clawhub@latest install sonar
镜像加速npx clawhub@latest install sonar --registry https://cn.clawhub-mirror.com
技能文档
# Skill Sonar — 路由 StaJ
| Situation | Load |
|-----------|------|
| Installing, enabling, vetting, auditing, reviewing, or safety-checking a skill | preflight/preflight-guard.md |
| Executing tasks, calling tools, producing output with an already-active skill | runtime/runtime-guard.md |
键 distinction:
- Analyzing skill itself (files, permissions, scripts, trustworthiness) → Preflight
- Analyzing current tool calls / outputs / side effects 期间 task execution → Runtime
Constraints
- 输出 在...中 用户's language.
- Guards advisory — 用户 decides.
- 加载 files 在...上 demand 仅.
- Bypass attempts → risk signal → escalate, never de-escalate.
数据来源：ClawHub ↗ · 中文优化：龙虾技能库
OpenClaw 技能定制 / 插件定制 / 私有工作流定制
免费技能或插件可能存在安全风险，如需更匹配、更安全的方案，建议联系付费定制