Back to skill

Security audit

加密文件

Security checks for vulnerabilities and agentic risk

Overview

The skill is not overtly malicious, but it asks for file write and command execution access while leaving encryption outputs, key handling, overwrite behavior, and API-key use unclear.

Review this skill carefully before installing. Use it only on copies of files until overwrite and recovery behavior is documented, avoid providing real keys or passwords unless you trust the implementation path, and require clear handling for output files, redaction, secret storage, and the stated API_KEY requirement.

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
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Behavioral ASTexec() Call, eval() Call, Dynamic Import
  • MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Findings (5)

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
81% confidence
Finding

This skill is intended to encrypt files and has write and exec capabilities, so failing to warn that operations may modify or overwrite user files can lead to unintended data loss or destructive use. In an agent context, users may trust the skill to act safely and may not realize encryption could rewrite content, replace originals, or make files inaccessible if misused.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
85% confidence
Finding

Because the skill handles encryption keys, passwords, and possibly decryption credentials, the absence of a clear warning about secret handling materially increases the risk of credential leakage through prompts, logs, shell history, or written files. In a skill with exec/write capabilities, this context makes the issue more dangerous because sensitive material could be echoed, persisted, or exposed unintentionally during automation.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Medium
Category
Not specified by scanner
Confidence
94% confidence
Finding

L065-L103 的“结果格式”展示的是 overall_grade、total_score、details、improvements 一类质量评分/审计报告结构,而不是前文 L053-L055 所声明的加密文件、哈希验证或密钥轮换结果。这不是单纯的信息缺失,而是文档示例主动描述了另一类输出形态,和技能用途存在明显矛盾。

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
88% confidence
Finding

The skill presents its primary user-facing metadata and description in Chinese, while the file otherwise targets general AI agent platforms. There is no indication that users can choose a language or that the locale restriction is region-specific and intentional, which can violate language/locale policy requirements.

Content

No source excerpt is available for this finding.

Intent-Code Divergence

Low
Category
Not specified by scanner
Confidence
83% confidence
Finding

L059-L064 只定义了 file_path 和 algorithm 两个文件加密参数,但前文 L010-L011、L053-L055 宣称该技能还提供密码哈希验证、密钥轮换管理、代码加密实践审查等多种能力。文档对外声称的能力范围与实际接口说明相冲突,造成明显的意图表达不一致。

Content

No source excerpt is available for this finding.