ClawHub
Back to skill

Security audit

Weighted Blanket Buying Consultant

Security checks for vulnerabilities and agentic risk

Overview

This is an instruction-only shopping guidance skill with disclosed product suggestions and no code, credentials, persistence, or hidden execution behavior.

Before installing, understand that this skill may ask for body weight, age, country, and health-related cautions to make safety-aware recommendations. Treat any named blanket models as examples to verify independently for current availability, specs, price, and suitability, not as medical advice or guaranteed impartial market coverage.

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 (1)

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The skill advertises 'no sales bias' and 'pure spec-first guidance,' but later instructs the agent to provide named product recommendations and even seeds a list of preferred reference models. That creates a deceptive trust signal: users may treat the output as impartial while receiving brand/model steering, which can be abused for covert promotion or undisclosed affiliate-style influence.

Content

No source excerpt is available for this finding.