ClawHub
Back to skill

Security audit

Iterative Code Review

Security checks across malware telemetry and agentic risk

Overview

This is a transparent code-review helper with optional automation that can edit code only when the user enables it, so it is acceptable with caution.

Install only if you want an iterative code-review workflow. Keep autoFix and autoContinue disabled for important repositories unless you are on a clean branch and are comfortable reviewing or reverting generated edits. Approve npm build/test execution only for repositories you trust or in a sandbox.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (3)

Vague Triggers

Medium
Confidence
89% confidence
Finding
The skill description uses very broad trigger terms like 'review', '审查', and '检查代码', which are common in ordinary coding conversations and can cause the skill to activate when the user did not specifically intend to invoke this workflow. In an agent setting, unintended invocation can expand the action surface to repository inspection, subagent spawning, and optional automated fix paths, increasing the chance of unnecessary or risky operations.

Natural-Language Policy Violations

Medium
Confidence
94% confidence
Finding
The pre-flight content begins in Chinese and presents the required review prompt only in Chinese, without offering a language choice or fallback. In a security-sensitive review workflow, forcing a single language can cause user misunderstanding of model selection and review-scope confirmations, increasing the chance of incorrect approvals or skipped checks.

Missing User Warnings

Medium
Confidence
90% confidence
Finding
The workflow allows automatic spawning of a Fixer based on a local preference file, without an explicit user-facing confirmation when `autoFix=true`. In a code-review skill, this expands behavior from analysis into code modification, which can cause unintended edits, alter repositories the user expected to remain read-only, and chain into further automated actions before the user notices.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.