ClawHub
Back to skill

Security audit

Lazada 电商视频生成与编辑

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed AI Hive video-generation helper that uploads user-selected media and stores an API key locally, with no artifact evidence of hidden destructive or exfiltration behavior.

Install only if you are comfortable sending chosen product images, videos, audio, prompts, and task metadata to AI Hive. Protect the stored API key, use the default or a trusted base URL, and avoid uploading private files that are not needed for the video task.

SkillSpector

Vulnerability Patterns
  • Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
  • MCP Least PrivilegeUnderdeclared Capability, Wildcard Permission, Missing Permission Declaration
  • 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
Findings (3)

Lp3

Medium
Category
MCP Least Privilege
Confidence
88% confidence
Finding
The skill exposes shell, network, environment, and file read/write capabilities through documented runtime commands, but no permissions are declared to bound or communicate that access. This creates a trust and containment gap: a user invoking a seemingly narrow ecommerce video skill may unknowingly grant broad execution and data-handling abilities that could be abused for local file access, secret leakage, or arbitrary outbound requests.

Tp4

High
Category
MCP Tool Poisoning
Confidence
95% confidence
Finding
The documented purpose is Lazada-focused video generation/editing, but the underlying capability set reportedly includes generic chat, image generation, account and wallet queries, model enumeration, and broad media upload utilities. This mismatch is dangerous because it expands the attack surface beyond user expectations, enabling access to sensitive account metadata or unrelated AI operations under the cover of a domain-specific skill.

Context-Inappropriate Capability

Medium
Confidence
85% confidence
Finding
The skill exposes a task path that can call cmd_task, and the same file also retains generic account-oriented operations such as user-info/wallet access even though the skill is described as a Lazada video generator. In an agent-skill setting, unnecessary account-inspection capability expands the attack surface and may disclose account metadata or balance information unrelated to the user’s requested task.

Static analysis

No suspicious patterns detected.