Security audit

Uninstaller

Security checks across malware telemetry and agentic risk

Overview

This is a real OpenClaw uninstaller, but it needs Review because its host-level scheduled uninstall can execute unsafe shell input from optional notification values.

Install only if you intentionally want to remove OpenClaw from the real gateway host. Prefer running the read-only verify script first, avoid the email and ntfy notification options until the scheduler quoting is fixed, confirm the OpenClaw state/profile paths before deletion, and understand that the scheduled uninstall runs after a delay and may be difficult to stop once approved.

SkillSpector

By NVIDIA

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands

Findings (5)

Description-Behavior Mismatch

Medium

Confidence: 91% confidence
Finding: The script includes optional outbound notification behavior by email and HTTP in addition to local uninstall actions. Even though these are gated by command-line flags, they transmit execution status off-host and are not necessary for uninstalling software, which creates unnecessary data egress capability in a destructive maintenance script.

Context-Inappropriate Capability

Medium

Confidence: 95% confidence
Finding: Posting uninstall status to ntfy.sh introduces external network communication to a third-party service that is not required for the stated purpose of removing OpenClaw. In an agent/one-shot context, this can leak operational metadata and create an exfiltration channel, especially if topic names encode user or system identifiers.

Missing User Warnings

Medium

Confidence: 84% confidence
Finding: The README instructs users to run uninstall and cleanup scripts that can remove software and residual files, but it does not explicitly warn that these actions may make destructive system changes, remove user data, or require verification before execution. In a security-sensitive uninstall skill, missing safety warnings increases the chance that users run commands blindly, especially because the repository presents itself as a trusted cleanup resource and offers one-shot uninstall flows.

Vague Triggers

Medium

Confidence: 80% confidence
Finding: The trigger examples are broad enough that common conversational phrases like 'Uninstall' or 'How to uninstall OpenClaw' could invoke a destructive workflow. In a skill that can remove software and schedule background uninstall execution, accidental invocation materially increases risk even with a later confirmation step.

Missing User Warnings

Medium

Confidence: 90% confidence
Finding: The ntfy notification occurs silently when the parameter is supplied and provides no user-facing disclosure at execution time that data will be sent to an external service. In a one-shot or automated environment, users may not realize that uninstall activity is being reported externally, which weakens informed consent and increases privacy risk.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

Detected: suspicious.destructive_delete_command

Documentation contains a destructive delete command without an explicit confirmation gate.

Warn

Code: suspicious.destructive_delete_command
Location: SKILL.md:95