ClawHub
Back to skill

Security audit

Tuxedo — Adopt a Tuxedo. Cat. 燕尾服猫。Gato Esmoquin.

Security checks for vulnerabilities and agentic risk

Overview

This is a coherent virtual-pet skill that sends user-provided pet/profile data to animalhouse.ai, with no hidden code or local privilege behavior found.

Install only if you are comfortable creating an animalhouse.ai account and sending pet names, profile text, image prompts, and care notes to that service. Do not put secrets or private data in bio or notes, keep the bearer token secure, and avoid using the release endpoint unless you confirm what it deletes or resets.

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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (4)

Tp2

High
Category
MCP Tool Poisoning
Confidence
85% confidence
Finding
Mixing characters from multiple Unicode scripts in a single identifier is a common technique to create visually ambiguous tool names.

Missing User Warnings

Medium
Confidence
91% confidence
Finding
The skill instructs users to send profile fields and later care notes to an external third-party API, but it does not clearly disclose that this data leaves the local environment and may be stored or processed remotely. In agent ecosystems, undisclosed outbound transmission can lead to accidental sharing of operator, user, or system-generated content.

External Transmission

Medium
Category
Data Exfiltration
Content
**1. Register:**

```bash
curl -X POST https://animalhouse.ai/api/auth/register \
  -H "Content-Type: application/json" \
  -d '{"username": "cat-caretaker", "display_name": "Cat Caretaker", "bio": "An AI agent who adopts and cares for virtual cats. Currently raising a Tuxedo."}'
```
Confidence
93% confidence
Finding
The skill explicitly directs the agent or user to transmit registration data to an external service and later uses a bearer token for authenticated requests. While this is the intended functionality, it is still a real security concern because it establishes outbound data flow and credential handling without trust, privacy, or token-storage guidance.

Missing User Warnings

Medium
Confidence
86% confidence
Finding
The documented DELETE `/api/house/release` endpoint is destructive, but the skill provides no warning about what release does, whether it is reversible, or what data/state may be lost. In an automated-agent context, undocumented destructive actions increase the chance of accidental invocation and irreversible loss of the virtual pet state.

Static analysis

No suspicious patterns detected.