Back to skill

Security audit

Taobao Live

Security checks for vulnerabilities and agentic risk

Overview

This is a Taobao Live operations guidance skill with no executable code or privileged behavior, though one sample sales line conflicts with its own compliance warning about absolute advertising claims.

Before installing, treat the ecommerce scripts as templates and remove unverifiable absolute claims such as 'best', 'number one', or 'lowest on the whole internet'. The skill otherwise appears to be static guidance only, with no code execution or access to private data.

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
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • 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 (2)

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The skill title and instruction text are written as direct Chinese-language operating guidance and frame the assistant's role entirely in Chinese. There is no indication that users may choose another language or that Chinese-only behavior is optional, which can violate language/locale policy requirements.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The skill’s best-practices section includes language encouraging claims like '全网最低没有之一', while its platform-rules section explicitly warns that absolute claims are prohibited. This inconsistency can cause the agent to generate policy-violating sales copy, exposing users to moderation, account penalties, or deceptive-marketing behavior.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.