ClawHub
Back to skill

Security audit

OpenClaw Dashboard

Security checks across malware telemetry and agentic risk

Overview

This appears to be a local OpenClaw dashboard, but it needs review because its installer creates persistent background services for missing helper code without clear opt-in.

Install only after reviewing or obtaining the complete dashboard source, especially server.py, autoheal.py, and app_watchdog.py. Prefer the foreground ./run_monitor.sh path first, avoid install_launchd.sh unless you explicitly want persistent launchd services, and use uninstall_launchd.sh to remove those services if already installed.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (3)

Description-Behavior Mismatch

Medium
Confidence
94% confidence
Finding
The installer creates and starts persistent background services for auto-heal and watchdog behavior, which goes beyond a simple local dashboard launcher and increases the software's authority and dwell time on the host. In the context of a local supervision dashboard, silently installing multiple always-on agents creates unnecessary persistence and broadens the attack surface if any of the referenced Python scripts are compromised or behave unexpectedly.

Context-Inappropriate Capability

Medium
Confidence
97% confidence
Finding
The script writes launch agent plists into ~/Library/LaunchAgents and enables automatic restart via RunAtLoad, KeepAlive, and scheduled relaunch behavior, establishing OS-level session persistence. Even though this is user-level rather than root-level persistence, it is still security-relevant because it causes code to execute automatically on login and can hide long-lived behavior from users expecting a local app launcher.

Missing User Warnings

Medium
Confidence
96% confidence
Finding
The script performs persistent filesystem writes and immediately installs/enables launch agents without any prior warning, confirmation, or dry-run mode. This is dangerous because users may believe they are starting a local dashboard while instead granting the skill durable background execution across future sessions.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.