Security audit

文档

Security checks for vulnerabilities and agentic risk

Overview

This skill is not plainly malicious, but it makes strong encryption and collaboration claims while requesting broad file and command access without enough implementation detail or limits.

Review carefully before installing. Do not rely on this skill for confidential or regulated documents unless the publisher supplies a real encryption design, key-management details, and clear limits on file access and command execution. Treat documents processed through it as potentially visible to the agent runtime and any tools the agent invokes.

A.I.G

Vulnerability Patterns

Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Insecure Skill Coding PracticesFinds exploitable flaws such as hardcoded secrets or command injection
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

Findings (2)

T05 · Unauthorized Access and Privilege Escalation

Warning

Location: SKILL.md:22
Finding: Excessive and Undefined Command Execution and Filesystem Permissions
Content: View full analysis
Remediation: View remediation

T09 · Insecure Skill Coding Practices

Warning

Location: SKILL.md:41
Finding: Unsupported End-to-End Encryption Security Claims
Content: View full analysis
Remediation: View remediation

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 (5)

Intent-Code Divergence

High

Category: Not specified by scanner
Confidence: 96% confidence
Finding: The skill is presented as encrypted collaborative document tooling, but then advertises unrelated capabilities such as vulnerability scanning, compliance auditing, asset risk scoring, and threat intelligence. This scope mismatch is dangerous because it obscures the true behavior and may justify broader permissions, external connectivity, or data handling than users expect from a document tool.
Content: No source excerpt is available for this finding.

Intent-Code Divergence

Medium

Category: Not specified by scanner
Confidence: 96% confidence
Finding: The skill claims end-to-end encrypted document handling, but the content only describes generic file processing, API usage, and command execution with no concrete encryption design, guarantees, or implementation boundaries. This is dangerous because users may trust the skill with sensitive documents under a false security assumption, causing plaintext exposure to the agent runtime, external APIs, logs, or shell commands.
Content: No source excerpt is available for this finding.

Intent-Code Divergence

Medium

Category: Not specified by scanner
Confidence: 91% confidence
Finding: The document says it is not for cracking encrypted files, yet it exposes broad exec and API-integrated behavior without encryption-specific safeguards, scope restrictions, or command limitations. That combination can mislead reviewers and users into assuming safety while still enabling arbitrary processing of sensitive files or misuse of command execution pathways.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 92% confidence
Finding: The invocation guidance is overly broad, telling the agent to use the skill for general file handling, document conversion, format conversion, and content extraction. Broad triggers increase the chance the agent invokes a high-permission skill unnecessarily, exposing user documents to read/write/exec behaviors outside a narrowly intended security context.
Content: No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium

Category: Not specified by scanner
Confidence: 81% confidence
Finding: The natural-language description states '支持中文交互，无需复杂配置即开即用', which presents Chinese-language interaction as the expected mode without offering a user choice. Language constraints should be optional or explicitly user-selectable unless clearly justified as region-specific.
Content: No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.