ClawHub
Back to skill

Security audit

code-review

Security checks across malware telemetry and agentic risk

Overview

This skill is a disclosed PR review workflow that uses GitHub and subagents in ways that match its stated purpose, with mutation steps mostly confirmation-gated.

Install this if you want a PR-review orchestrator that may use your GitHub CLI session to request Copilot review and coordinate subagents. Be aware that saying ordinary phrases like “review this PR” can invoke it, and confirm carefully before any push or issue creation.

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 (1)

Vague Triggers

Medium
Confidence
95% confidence
Finding
The skill explicitly states that it is deliberately model-invocable and uses broad natural-language triggers such as "review this PR." Everyday phrasing can cause accidental invocation in unrelated contexts, which is risky because this skill performs external-state actions like requesting Copilot review and potentially editing PR state after confirmation-gated steps. In a code-review orchestration skill, unintended activation can waste resources, trigger unwanted GitHub operations, or steer an agent into a more privileged workflow than the user intended.

VirusTotal

64/64 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.