ClawHub
Back to skill

Security audit

pet-food-compare

Security checks for vulnerabilities and agentic risk

Overview

This skill is a transparent pet-food comparison guide with no code execution, persistence, or hidden data access, though users should treat its pet-health mappings as informational rather than veterinary advice.

Installers should understand that this skill performs public web/image/OCR research and summarizes product facts. Do not treat ingredient-to-condition mappings as veterinary recommendations, especially for prescription diets, urinary issues, allergies, obesity, kidney disease, or other health conditions.

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
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • 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)

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The skill claims to be a neutral product-comparison tool, but these instructions extend into mapping ingredient and nutrient patterns to animal health conditions. Even with wording like '对应/关联' and anti-recommendation disclaimers, this steers the model toward quasi-medical suitability guidance for pets, which users may treat as actionable health advice. In a pet food context, that can influence feeding decisions for animals with obesity, allergies, urinary issues, or other conditions without veterinary oversight.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

The document explicitly requires that the entire output be in Simplified Chinese and reiterates that the skill's primary users are mainland China users. This is a natural-language locale policy constraint applied unconditionally, and the file does not offer a user language choice or opt-in mechanism.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Low
Category
Not specified by scanner
Confidence
96% confidence
Finding

Line L249 instructs the agent to use an HTML span with inline color styling for the '对应动物情况' field in HTML-rendered output. Later documentation explicitly says not to use HTML tags such as '' because most clients do not render them and they harm readability. These instructions conflict and create intent-code/documentation divergence within the skill guidance.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.