ClawHub
Back to skill

Security audit

Academic Thesis Workflow

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed academic writing workflow that uses bundled Markdown references and does not request system access, persistence, credentials, or network execution.

Installers should expect a Chinese-first academic drafting assistant. Specify a different language if needed, use it in line with academic integrity rules, and independently verify citations and generated scholarly claims before relying on them.

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
  • 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
Findings (3)

Natural-Language Policy Violations

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

The embedded academic prompt hard-codes the output language as Chinese ('语言为{中文}') without indicating that this is optional or user-selectable. This can override user preferences or system expectations, creating prompt-level behavior that may mis-handle requests, produce unusable outputs for non-Chinese users, and interfere with downstream review, safety, or compliance processes that assume the user's chosen language.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
93% confidence
Finding

The skill description is scoped so broadly ('supports all disciplines' and should be used whenever a user wants to write a thesis, dissertation, or paper) that it may activate for a wide range of ordinary academic-writing requests. Over-broad invocation can cause unintended routing, reducing user control and potentially steering users into a deterministic paper-generation workflow even when they only asked for limited editing, brainstorming, or research assistance.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The skill hard-codes the output language as Chinese in the prompt template ("语言为{中文}") instead of respecting the user's language preference. This can override user intent, reduce usability, and in multilingual environments may cause the agent to ignore higher-priority runtime preferences or accessibility needs.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.