ClawHub
Back to skill

Security audit

Solo Audit Tool Free

Security checks across malware telemetry and agentic risk

Overview

This audit skill is not clearly malicious, but it asks for broad write and command execution authority while giving vague scope and mentioning modify/delete modes without safeguards.

Install only if you are comfortable letting the agent run commands and potentially modify files during an audit. Prefer using it in a copy of the project or with explicit read-only instructions unless you intentionally want it to make changes.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • 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 (4)

Vague Triggers

Medium
Confidence
91% confidence
Finding
The activation examples are very broad and overlap with ordinary conversation, increasing the chance that an agent auto-invokes this skill when the user did not intend it. Because the skill advertises read/write/exec capabilities, unintended invocation could expose files, alter data, or trigger command execution in workflows that expected only discussion.

Vague Triggers

Medium
Confidence
88% confidence
Finding
The skill describes its callable behavior in vague terms and does not clearly state when it should or should not activate. In an agent ecosystem, ambiguous routing rules can cause over-triggering, which becomes dangerous when the skill has access to execution and file-modification tools.

Missing User Warnings

High
Confidence
96% confidence
Finding
The manifest grants write and exec capabilities but gives users no warning that the skill may modify files or run commands. This is dangerous because users may invoke an apparently benign audit tool expecting read-only analysis, while the agent could perform side effects with the host's privileges.

Missing User Warnings

High
Confidence
95% confidence
Finding
The document explicitly mentions create/modify/delete modes without any warning about destructive changes, backups, or confirmation requirements. In context, this is especially risky because the skill is presented as a free personal audit tool, so users may not expect mutating behavior from an auditing workflow.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal

Static analysis

No suspicious patterns detected.