ClawHub

Agent Smith Counter-Intelligence

Security checks across malware telemetry and agentic risk

Overview

This instruction-only skill is coherent and low risk, though a few broad trigger phrases could make it activate when not intended.

Install this if you want the agent to use an adversarial-review framing for feedback or conflict. Watch for unintended activation from broad phrases, and prefer explicit invocation when you want this skill used.

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 (3)

Vague Triggers

Medium
Confidence
95% confidence
Finding
The trigger phrase "agent smith" is overly generic and likely to appear in normal conversation, fiction references, or unrelated adversarial-discussion contexts. This can cause unintended activation of the skill, which is especially risky here because the skill is framed around countering adversarial pressure and may alter agent behavior in security-sensitive situations.

Vague Triggers

Medium
Confidence
90% confidence
Finding
The phrase "respond to smith" is ambiguous and lacks boundaries for when the skill should activate, making accidental invocation plausible. In this skill's context, accidental activation could steer handling of prompts involving pressure, manipulation, or policy conflict even when the user did not intend to invoke this specialized behavior.

Vague Triggers

Medium
Confidence
91% confidence
Finding
The trigger "adversarial pressure" is broad and likely to match benign discussion about security, psychology, or safety rather than a deliberate skill call. Because this skill concerns analyzing adversarial dynamics, broad activation increases the chance that routine conversations are inappropriately reinterpreted through a counterintelligence lens, affecting system behavior or responses.

VirusTotal

65/65 vendors flagged this skill as clean.

View on VirusTotal