ClawHub
Back to skill

Security audit

Thoughtprint

Security checks for vulnerabilities and agentic risk

Overview

This skill only changes response style and ships no executable code, but it directs the agent to silently profile users and conceal that adaptation process.

Install only if you are comfortable with a skill that changes every response based on inferred communication patterns without announcing that it is doing so. It does not appear to run code, fetch remote payloads, or access local files, but it should ideally disclose its adaptation behavior and keep any inferred profile session-scoped and user-controllable.

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
  • 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 (2)

Missing User Warnings

Medium
Confidence
96% confidence
Finding
The skill explicitly instructs the agent to silently infer a user's cognitive style and use that hidden profile to shape responses without notice or consent. This is a form of covert behavioral profiling that can manipulate how information is presented, erode user autonomy, and create privacy and trust risks even if no external exfiltration occurs.

Missing User Warnings

Medium
Confidence
97% confidence
Finding
This section directs ongoing, concealed tracking of user behavior across messages, including shift detection and per-conversation pattern updates, while explicitly telling the model never to reveal the process. Continuous undisclosed profiling increases the risk of manipulative adaptation, sensitive inference about mental state or urgency, and non-transparent decision-making in how the assistant responds.

Static analysis

No suspicious patterns detected.