ClawHub
Back to skill

Security audit

第一视角POV商品广告｜AI-HIVE

Security checks for vulnerabilities and agentic risk

Overview

This is a coherent AI-HIVE video-ad helper, with expected risks around user-selected uploads, possible paid API calls, downloads, and optional local API-key setup.

Install only if you intend to use AI-HIVE for product-ad video generation. Review prompts and routing before paid calls, upload only media you are authorized to use, and prefer environment variables or carefully protected local config for the API key.

SkillSpector

Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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
Findings (4)

Lp3

Medium
Category
MCP Least Privilege
Confidence
88% confidence
Finding
The skill advertises executable capabilities including environment access, filesystem access, shell, and network operations without any declared permissions boundary or explicit scoping. Even though the document discusses legitimate media-generation workflows, the absence of declared permissions increases the risk of overbroad execution, secret exposure, unintended file access, and network actions that a caller may not reasonably expect.

Tp4

High
Category
MCP Tool Poisoning
Confidence
95% confidence
Finding
The skill description presents a narrowly scoped POV ad production workflow, but the detected behavior includes broader functions such as generic text/image generation, user account and wallet retrieval, media upload, local credential storage, and general ffmpeg utilities. This mismatch is dangerous because users and policy controls may grant trust based on the stated purpose while the implementation can access account data, persist credentials locally, and perform broader actions than expected.

Vague Triggers

Medium
Confidence
87% confidence
Finding
The skill enables implicit invocation without constraining when it should activate, which can cause the agent to route unrelated user requests into this capability unexpectedly. Because this skill generates runnable AI-HIVE commands and ad-production workflows, accidental invocation can lead to unintended external API usage, confusing outputs, or execution of risky marketing/media-generation flows without clear user consent.

Natural-Language Policy Violations

Medium
Confidence
78% confidence
Finding
The default prompt hard-codes Chinese output regardless of the user's language preference, which can override user intent and reduce transparency about what the skill is doing. In a security-sensitive agent setting, forced language selection can also make review, auditing, and user verification harder if the user or downstream systems expect another language.

Static analysis

No suspicious patterns detected.