ClawHub
Back to skill

Security audit

S.H.I.T Journal 论文推荐

Security checks for vulnerabilities and agentic risk

Overview

This is a simple satirical paper-recommendation skill that opens one disclosed website, with no bundled code, persistence, or privileged access.

Install this only if you want an agent to browse S.H.I.T Journal for satirical paper recommendations. Be aware that broad prompts like “random paper” may invoke it, and the target site may present 18+ content or consent gating.

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
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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)

Vague Triggers

Medium
Confidence
94% confidence
Finding
The skill metadata declares broad trigger phrases for a lightweight content-recommendation action, including generic terms like '随机论文' that could match ordinary user requests unrelated to this specific satire site. This can cause unintended invocation and context confusion, especially because the skill performs browser-based navigation to an external site.

Vague Triggers

Medium
Confidence
94% confidence
Finding
The invocation section lists ambiguous examples in both Chinese and English without boundaries, making it likely that normal academic or humorous-paper queries could accidentally trigger this skill. Because the skill then browses an adult-gated external website, accidental activation increases exposure to unexpected content and unnecessary external requests.

Static analysis

No suspicious patterns detected.