ClawHub
Back to skill

Security audit

Qqlive

Security checks for vulnerabilities and agentic risk

Overview

This is a plain Tencent Video guidance prompt with no code, persistence, or privileged access, though its advice scope is broader than its description.

Read the broader content and creator-advice sections before installing if you only wanted membership and device-limit troubleshooting; verify Tencent Video prices and rights on the official purchase or content pages because those details change frequently.

A.I.G

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

SkillSpector

Vulnerability Patterns
  • 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
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
92% confidence
Finding

The skill internally frames itself as an AI content assistant for recommendations, analysis, and creator guidance, which contradicts the manifest's much narrower operational purpose. This mismatch is risky because orchestration, permissioning, and user expectations may rely on the manifest, while the actual instructions steer the model into broader advisory behavior not reviewed for that use case.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The manifest promises a narrow Tencent Video guide focused on membership tiers, content-rights mapping, and device/download troubleshooting, but the body broadens into general entertainment recommendation, plot analysis, review writing, and creator monetization advice. This scope drift is dangerous because downstream systems or users may grant or trust the skill for a constrained purpose while it actually influences unrelated domains, increasing the chance of policy bypass, user deception, or unintended use.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.