Back to skill

Security audit

华山论卷

Security checks for vulnerabilities and agentic risk

Overview

This is a Chinese-language teaching assistant skill for reviewing and generating school exam-paper materials, with no code, persistence, credential use, or hidden actions found.

Installers should expect this skill to respond in a Chinese teaching-research style and to activate around exam-paper, question-writing, and assessment discussions. It appears low risk from a security perspective, but users should still avoid sharing confidential student data or copyrighted real exam materials unless they have permission.

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
Vulnerability Patterns
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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
Category
Not specified by scanner
Confidence
95% confidence
Finding

The trigger scope is broad because it activates on generic education terms like '试卷', '出题', '组卷', '考题', '命题', and '考情', which can appear in many benign conversations beyond the intended workflow. This can cause unintended activation, context hijacking, or inappropriate insertion of the skill into unrelated chats, especially in multi-skill routing systems where overbroad matching increases misfires.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The instruction at L012 fixes the assistant’s tone and style in Chinese and the overall skill content assumes a single language/locale, but there is no explicit user opt-in or alternative language option. Under the policy, forcing a specific language without choice can be a natural-language policy violation unless the locale restriction is clearly documented and justified.

Content

No source excerpt is available for this finding.