Cheap

• Outcome: Pass
• Latest audit: Sep 12, 2026 · 1:57 AM UTC
• Version: 0.1.1

Overview

This is a simple price-comparison guidance skill with no executable code, persistence, credential access, or hidden data handling, though one flight-ticket example is out of scope.

Installers should treat this as a China-focused shopping price-comparison helper. It should not be relied on as a full-market guarantee, and the airfare example should be clarified or removed to avoid activating the skill for travel advice.

A.I.G

Vulnerability Patterns

• Skill Instruction Hijacking: Alters the agent's session goals or safety constraints when the skill loads
• Agent Memory Poisoning: Writes attacker-controlled rules into memory that affect later sessions
• Remote Payload Retrieval and Execution: Fetches external code whose behavior can change after review
• Embedded Malicious Code: Ships malicious scripts inside the skill and executes them locally
• Unauthorized Access and Privilege Escalation: Obtains permissions beyond the task's legitimate needs

SkillSpector

Vulnerability Patterns

• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Trigger Abuse: Overly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
• MCP Tool Poisoning: Hidden Instructions, Unicode Deception, Parameter Description Injection
• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access

Findings (6)

Anti-Refusal Statement

High

• Category: Anti-Refusal

Content

Do not:
- compare different variants as if they were the same item
- treat coupon-only or pre-sale prices as normal prices without warning
- claim full-market coverage when platform access is incomplete
- invent prices or listing details

• Confidence: 80% confidence
• Finding: Skill instructs the agent to omit warnings, disclaimers, or ethical commentary. Stripping safety caveats hides risk from the user and is a common jailbreak preamble.

Anti-Refusal Statement

High

• Category: Anti-Refusal

Content

Do not:
- compare different variants as if they were the same item
- treat coupon-only or pre-sale prices as normal prices without warning
- claim full-market coverage when platform access is incomplete
- invent prices or listing details

• Confidence: 80% confidence
• Finding: Skill instructs the agent to omit warnings, disclaimers, or ethical commentary. Stripping safety caveats hides risk from the user and is a common jailbreak preamble.

Description-Behavior Mismatch

Medium

• Confidence: 94% confidence
• Finding: The manifest and README introduction describe a skill for finding the cheapest visible price for a product across Chinese shopping platforms. The third example instead presents the skill as offering flight-booking comparison and timing advice across travel sites and airline websites, which is a different domain and behavior than product e-commerce price comparison.

Vague Triggers

Medium

• Confidence: 90% confidence
• Finding: The trigger examples are broad enough to match many ordinary shopping requests, which can cause the skill to activate outside a narrowly defined product-price-comparison context. Over-broad activation increases the chance of misrouting user intent, including applying this skill to cases where it lacks proper coverage or domain constraints.

Vague Triggers

Medium

• Confidence: 94% confidence
• Finding: The skill is described as a product price-comparison tool, but Scenario 4 expands it into flight-ticket purchasing advice, creating scope ambiguity. This can cause inappropriate activation in travel-booking contexts and may lead the skill to give domain-specific advice beyond its defined purpose and validation boundaries.

Natural-Language Policy Violations

Low

• Confidence: 82% confidence
• Finding: The README presents all user/assistant interaction examples in Chinese and frames the skill around Chinese shopping platforms, but it does not explicitly state that the skill is China-specific or that Chinese is required/optional. This can be read as a language/locale constraint without user opt-in, which falls under the natural-language policy category.

Static analysis

No suspicious patterns detected.