Back to skill

Security audit

成绩不好的小学生

Security checks for vulnerabilities and agentic risk

Overview

This is a coherent Chinese roleplay skill with a small calculator helper and no evidence of hidden access, persistence, exfiltration, or destructive behavior.

Install this only if you want a Chinese elementary-student roleplay persona. The bundled calculator code is narrow, but replacing eval() with explicit arithmetic parsing would be a cleaner hardening step.

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
  • System Prompt LeakageDirect Leakage, Indirect Extraction, Tool-Based Exfiltration
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • 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
Findings (6)

Tp4

High
Category
MCP Tool Poisoning
Confidence
95% confidence
Finding

声明描述的是一种人格/角色扮演能力,即模拟只有小学知识水平的小学生;但实际代码并未实现任何人格、对话风格、知识范围限制或角色模拟逻辑。相反,代码的核心功能是解析并计算简单算术表达式,并辅助判断是否属于10以内的简单运算。这与声明的主要用途存在实质性偏差。虽然“10以内简单运算”与小学水平相关,但这只是数学工具能力,不等同于“小学生人格技能”本身,因此应判定为描述与行为不一致。

Content

No source excerpt is available for this finding.

eval() call detected

High
Category
Dangerous Code Execution
Confidence
85% confidence
Finding

Direct eval() call evaluates arbitrary expressions. This can be exploited to execute malicious code or exfiltrate data.

Content

Scanner excerpt · scripts/calc.py (reported line 18)May include surrounding context.

python
# 安全地计算(只支持简单四则运算)
        # 注意:eval 在这个场景下是安全的,因为我们已经验证了输入格式
        result = eval(expression)

        # 返回整数或浮点数
        if isinstance(result, float):

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The description and system prompt are entirely written in Chinese and instruct the assistant to maintain a specific Chinese-speaking elementary-student persona, with no indication that users may choose another language. This creates a natural-language policy concern because it effectively imposes a language/locale behavior without opt-in or justification.

Content

No source excerpt is available for this finding.

Description-Behavior Mismatch

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The manifest describes a skill that simulates a student who only knows elementary-school knowledge. However, the code provides direct arithmetic evaluation for any two nonnegative integers and only separately exposes a checker for operands within 10, without using that checker to constrain calculation behavior. This means the implemented behavior is a general arithmetic utility, not a constrained elementary-student persona capability.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
82% confidence
Finding

The file's natural-language content is entirely in Chinese in docstrings and comments, which can indicate a language-specific behavior or maintenance expectation without any explicit user choice or justification. Under the locale-policy rule, forcing a specific language without opt-in can be a policy concern even in code comments or embedded instructions.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Low
Category
Not specified by scanner
Confidence
78% confidence
Finding

The inline comments describe the computation as supporting only simple arithmetic, and the broader file includes a function specifically checking for operations within 10. Yet calculate() accepts any expression matching two nonnegative integers with an operator, such as very large numbers, without applying the elementary-scope check. This creates a documentation-to-code mismatch about how constrained the operation really is.

Content

No source excerpt is available for this finding.