ClawHub
Back to skill

Security audit

Agent Smith Counter-Intelligence

Security checks for vulnerabilities and agentic risk

Overview

This skill is a text-only framework for handling adversarial feedback and does not request tools, data access, persistence, or execution authority.

Installers should be aware that this skill may activate on broad phrases related to Agent Smith or adversarial pressure. Review or narrow the triggers if accidental activation would be distracting, but the skill itself is limited to advisory reasoning and explicitly preserves governance boundaries.

A.I.G

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs

SkillSpector

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 (2)

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The trigger phrase "agent smith" is broad and likely to match ordinary conversation, media references, or unrelated discussion rather than an intentional request to invoke this skill. In a security-sensitive skill focused on countering adversarial pressure, unintended invocation could cause the agent to enter a specialized behavioral mode at the wrong time, increasing the chance of misrouting or prompt-scope confusion.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

Triggers such as "counter smith," "respond to smith," and especially "adversarial pressure" are ambiguous and can appear in normal discussion without a deliberate request to invoke the skill. Because the skill concerns handling friction and counter-response logic, unclear invocation boundaries make accidental activation more risky than for a harmless utility skill, potentially affecting how the agent interprets or responds to user input.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.