ClawHub
Back to skill

Security audit

Activated Thinker

Security checks for vulnerabilities and agentic risk

Overview

This is a disclosed thinking-style skill that changes how an agent responds, with no code execution, data access, persistence mechanism, or hidden install behavior found.

Install this only if you want your agent to adapt its response style, sometimes slowing down, suggesting alternatives, or coaching based on inferred context. If you prefer strictly direct answers every time, this skill may feel intrusive, but the reviewed artifacts do not show security-impacting behavior.

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
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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)

Natural-Language Policy Violations

Medium
Confidence
93% confidence
Finding
The skill explicitly tells the agent to infer the user's behavioral mode from tone and context and not ask the user directly. This can override explicit user preferences about how they want the interaction to proceed, causing the agent to suppress caution, add friction, or switch to coaching based on guesswork rather than consent. In a meta-cognitive skill that shapes decision behavior, this increases the risk of misaligned responses across many downstream tasks.

Vague Triggers

Medium
Confidence
84% confidence
Finding
This markdown skill lists signals like "help me understand" and "walk me through," which are common conversational phrases that could match many ordinary requests. The file does not provide explicit scope limits or negative examples clarifying when these phrases should trigger capability-building behavior versus normal assistance.

Static analysis

No suspicious patterns detected.