ClawHub
Back to skill

Security audit

代码质量审查

Security checks across malware telemetry and agentic risk

Overview

The skill appears to be a broad code-review helper, and the only supported concerns are scope clarity issues rather than hidden or harmful behavior.

Install this as a general review aid, not as an automatic gate for every small task. Prefer invoking it when you actually want a code or change review, and avoid feeding it sensitive repository context unless review is needed.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Vague Triggers

Medium
Confidence
87% confidence
Finding
The metadata-level trigger description is overly broad and lacks guardrails, which can cause this skill to be invoked in situations where a focused review or narrower tool would be more appropriate. In an agent ecosystem, overbroad activation increases unnecessary exposure to code, prompts, and potentially sensitive context, and can amplify misuse through habitual over-calling.

Vague Triggers

Medium
Confidence
91% confidence
Finding
The usage guidance says to use the skill before merging any PR or change, after any bug fix, and in many other broad cases, but provides no limiting conditions or exception handling. That creates a policy-level overreach vulnerability: the skill may become a mandatory catch-all reviewer, increasing prompt surface area, overcollection of context, and the chance that the agent performs security review outside intended trust boundaries.

VirusTotal

58/58 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.