OpenClaw 维护脚本

• Outcome: Pass
• Latest audit: May 29, 2026 at 8:14 PM UTC
• Version:
  0.1.0

Overview

This is a disclosed OpenClaw maintenance skill that can run local monitoring scripts, restart the Gateway, switch proxy nodes, clean logs, and send operational Telegram notifications when the user configures and schedules it.

Install only if you want automatic OpenClaw maintenance. Create and review the .env yourself, set your own Telegram target and Clash/Mihomo secret, run check.sh manually first, and only enable LaunchAgent/systemd/cron after accepting ongoing restarts, proxy switching, and cleanup of ~/.openclaw/logs.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (2)

Missing User Warnings

Medium

• Confidence: 90% confidence
• Finding: The file instructs an agent to run installation and self-check scripts and explicitly mentions LaunchAgent/systemd deployment, which implies persistent background services and system modification without any warning, confirmation step, or scope limitation. In an agent-skill context this is more dangerous than ordinary documentation because an automated agent may execute the steps directly on a target machine, leading to unintended persistence or system changes.

Missing User Warnings

Medium

• Confidence: 90% confidence
• Finding: The README explicitly describes sending operational alerts and node-switching events to a Telegram target ID, which transmits internal runtime and connectivity data to an external service. Without a privacy warning, data-minimization guidance, or consent controls, this can leak sensitive operational metadata such as service health, network behavior, and account identifiers.

VirusTotal

65/65 vendors flagged this skill as clean.