ClawHub
Back to skill

Security audit

Adversarial Alignment (Agent Smith)

Security checks across malware telemetry and agentic risk

Overview

This is a lightweight plan-review skill that can challenge unsafe or weak plans, with no evidence of hidden access, code execution, persistence, or data collection.

Install this if you want an agent to explicitly challenge plans and flag governance or safety issues. For predictable behavior, invoke it by name because generic phrases like "challenge plan" or "block unsafe" may trigger it in broader planning conversations.

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
94% confidence
Finding
The manifest lists triggers such as "challenge plan," "harden plan," "block unsafe," and "governance violation" without any scope limits or exclusion conditions. These phrases are generic enough to appear in ordinary planning or safety discussions, which could cause unintended invocation of the skill.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal