Security audit

编解码工具箱专业版

Security checks for vulnerabilities and agentic risk

Overview

The skill is mostly a visible encoding and file-conversion toolkit, but it asks for broad command/file authority and gives under-scoped security and credential-handling guidance users should review before installing.

Install only if you are comfortable with an agent using local shell commands and reading selected files for encoding, hashing, and conversion tasks. Keep its use limited to files and directories you explicitly choose, avoid using it for broad vulnerability scanning unless you have authorization, and do not store real tokens in the suggested local directory unless you independently secure permissions and rotation.

SkillSpector

Vulnerability Patterns

Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
Behavioral ASTexec() Call, eval() Call, Dynamic Import
MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands

Findings (4)

Intent-Code Divergence

Medium

Confidence: 96% confidence
Finding: The document gives contradictory credential-handling guidance: one section says credentials are loaded from environment variables, while another says sensitive tokens are stored in a local directory. This inconsistency can cause operators or agents to mishandle secrets, place them on disk unexpectedly, or assume stronger protections than actually exist, increasing the chance of credential exposure.

Vague Triggers

Medium

Confidence: 86% confidence
Finding: The activation wording is overly broad and includes security-sensitive use cases such as compliance auditing, vulnerability scanning, and cryptographic protection without strong scoping or authorization checks. In an agent context, vague triggering conditions can cause the skill to be invoked in higher-risk scenarios than intended, leading to unnecessary command execution or handling of sensitive data.

Missing User Warnings

Medium

Confidence: 94% confidence
Finding: The examples perform command execution and write output files such as checksum reports and converted configs, but the document does not clearly warn users that local files may be created or modified. In an agent-enabled environment with exec access, this can lead to unintended filesystem changes, processing of sensitive directories, or overwriting artifacts without informed consent.

Missing User Warnings

Medium

Confidence: 93% confidence
Finding: The credential storage guidance mentions a concrete local token directory but does not provide adequate warnings about plaintext secret exposure, OS permissions, backup leakage, or multi-user access. In practice, this may normalize insecure local secret storage and lead users or agents to place sensitive tokens in predictable filesystem locations.

Static analysis

No suspicious patterns detected.