ClawHub
Back to skill

Security audit

Pillar Alignment Check

Security checks across malware telemetry and agentic risk

Overview

This is a markdown-only content review checklist for Redditech Labs posts, with no executable code or hidden access behavior found.

Install this if you want an agent to enforce Redditech Labs-specific content pillar and brand voice rules. Before using the Outsider Perspective pillar, make sure any Caribbean, Australian, or personal-angle framing is user-approved and factually present in the content rather than inferred.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • 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
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
Findings (1)

Natural-Language Policy Violations

Medium
Confidence
94% confidence
Finding
The skill explicitly directs content to use specific geographic and cultural identity framing ("Caribbean. Australian. Building anyway.") and prescribes how that identity should be presented. If applied without the user's explicit consent or relevance to the source material, this can cause identity injection, misrepresentation, or disclosure of sensitive personal attributes in published content. In a content-review skill, this is more concerning because it can systematically pressure agents to add protected or personal identity cues to outward-facing posts.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal

Static analysis

No suspicious patterns detected.