ClawHub
Back to skill

Security audit

Concert

Security checks for vulnerabilities and agentic risk

Overview

This skill appears non-executable, but its stated concert-ticketing purpose conflicts with instructions that describe a video streaming product.

Review this skill carefully before installing because its advertised purpose and actual instructions do not match. It does not appear to run code or request privileges, but it may route requests incorrectly or produce irrelevant guidance until the publisher aligns the description and body content.

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 (3)

Description-Behavior Mismatch

High
Confidence
98% confidence
Finding
The declared skill description says it provides official ticket booking channels and venue check-in guides for concerts and stage performances, but the body documents a media streaming product. This broad scope mismatch is more serious than a simple wording issue because it can cause trust violations, incorrect invocation by an agent, and potential redirection of users to unintended functions or data flows under false pretenses.

Natural-Language Policy Violations

Medium
Confidence
92% confidence
Finding
The natural-language content of the skill is entirely in Chinese, including the description, headings, and example requests, with no indication that users may interact in other languages or opt into a Chinese-only experience. Under the policy for language/locale constraints, this is a potential forced-language behavior unless the restriction is explicitly documented and justified.

Intent-Code Divergence

Medium
Confidence
96% confidence
Finding
The document section titled '演出订票' claims to support concert and performance ticketing, but the actual fields and filters describe video streaming features such as 4K playback, bullet comments, playback speed, and youth mode. This mismatch can mislead users and downstream agents about the skill’s real capability, increasing the risk of unsafe routing, deceptive behavior, or misuse of unrelated systems.

Static analysis

No suspicious patterns detected.