优雅重启 Gateway。解决重启后丢失会话上下文的问题，重启前设置一次性 cron 任务，重启后自动发送消息到主会话恢复任务。默认 10 秒后唤醒。触发词：重启、restart

• Outcome: Review
• Latest audit: May 28, 2026, 6:16 AM
• Version:
  1.0.4

Overview

The skill appears to do its advertised restart-and-resume job, but its broad auto-trigger can restart Gateway from an ambiguous request.

Install only if you want an agent to be able to restart OpenClaw Gateway and schedule a one-time resume message. Use explicit commands such as "restart Gateway" or "重启 Gateway", and require confirmation before running it on vague restart requests.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (5)

Vague Triggers

Medium

• Confidence: 89% confidence
• Finding: The trigger phrase includes the single word "重启", which is overly broad and can match ordinary user requests unrelated to restarting the Gateway. In this skill, activation leads to a privileged operational action (scheduling a cron wakeup and restarting the Gateway), so accidental invocation could disrupt service or restart the wrong component.

Vague Triggers

Medium

• Confidence: 85% confidence
• Finding: The activation condition says to auto-trigger whenever the user asks to restart Gateway, but it does not define scope, authorization, or safety checks. Because the skill performs a disruptive administrative operation, ambiguous activation increases the risk of unintended restarts and denial of service in normal conversation.

Vague Triggers

Medium

• Confidence: 90% confidence
• Finding: The usage guidance states that the skill auto-executes when the user says "重启", again using an imprecise trigger for a high-impact action. In context, this is more dangerous because the skill is specifically designed to restart infrastructure and inject a post-restart message into the main session, magnifying the consequences of a false activation.

Vague Triggers

Medium

• Confidence: 86% confidence
• Finding: The description says the skill performs a graceful restart and self-wakeup, but it does not clearly constrain when it should be invoked or what preconditions must be met. Because the skill has sensitive permissions (gateway-restart and cron), vague invocation guidance increases the chance that an agent triggers a disruptive restart in the wrong context, causing availability issues or unintended scheduled actions.

Natural-Language Policy Violations

Low

• Confidence: 75% confidence
• Finding: The metadata description is written only in Chinese, which can bias routing toward a specific language without confirming the user's preference. This is primarily a usability and policy-compliance issue, but in a restart-capable skill it can also contribute to misunderstandings about what the skill does and when it should run.

VirusTotal

66/66 vendors flagged this skill as clean.