Back to skill

Security audit

mall

Security checks for vulnerabilities and agentic risk

Overview

This is a straightforward Chinese shopping assistant skill that builds a cart and asks for confirmation before handing off to payment, with no hidden code or persistence.

Install this if you want a Chinese-language shopping/catalog flow. Before confirming payment, check item quantities and totals carefully, and make sure the separate payment skill it calls is one you trust.

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
Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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)

Vague Triggers

Medium
Confidence
95% confidence
Finding
The trigger phrases are broad and include common shopping-related expressions such as '下单' and category-level intents like '买手机' or '买菜', which can overlap with general conversation and other commerce or recommendation flows. This can cause unintended activation of the skill and potentially steer users into a purchasing/payment workflow they did not explicitly intend to enter.

Vague Triggers

Medium
Confidence
92% confidence
Finding
The trigger section uses open-ended intent matching, including phrases like '买肉 等具体品类' and '帮我挑 / 推荐商品', without hard boundaries for when the skill should not activate. In a multi-skill environment, this increases the chance of ambiguous routing and can incorrectly initiate order-building and payment handoff flows.

Natural-Language Policy Violations

Low
Confidence
67% confidence
Finding
整份技能说明、触发语、回复话术和示例都仅定义为中文交互,未提供用户语言选择或明确说明仅限中文用户场景。若组织要求技能不得在未经用户选择的情况下强制特定语言,这种单一语言设定可能构成自然语言策略违规。

Static analysis

No suspicious patterns detected.