Security audit

文档扫描增强

Security checks for vulnerabilities and agentic risk

Overview

This skill is mainly a document/image enhancement guide, but it has mismatched scope claims and promotes risky alteration uses such as removing exam answers and watermarks.

Review this before installing. Use it only for images you are authorized to process, avoid sensitive personal or business documents unless you accept sending them to the external service, and do not use it to remove ownership marks, watermarks, or answers from records in a misleading way. The publisher should narrow the trigger text and remove unrelated security/coding claims.

A.I.G

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

SkillSpector

Vulnerability Patterns

Excessive AgencyUnrestricted Tool Access, Autonomous Decision Making, Scope Creep
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

Findings (5)

Vague Triggers

High

Category: Not specified by scanner
Confidence: 95% confidence
Finding: The description says the skill should be used for generic code generation, programming assistance, debugging, and deployment, even though the skill’s actual purpose is image/document enhancement. In an agent environment, this overbroad trigger can cause the skill to be selected in unrelated contexts and reach exec/read/write capabilities for tasks outside its intended scope, expanding attack surface and enabling prompt-routing abuse.
Content: No source excerpt is available for this finding.

Intent-Code Divergence

Medium

Category: Not specified by scanner
Confidence: 95% confidence
Finding: The paid-feature section introduces unrelated security capabilities such as vulnerability scanning, CVE correlation, and threat intelligence in a skill otherwise described as single-image document enhancement. This mismatch can mislead an agent or user about the skill’s scope, increasing the chance of inappropriate invocation, over-privileged trust, or accidental routing of sensitive security tasks to an unrelated tool.
Content: No source excerpt is available for this finding.

Ssd 4

Medium

Category: Not specified by scanner
Confidence: 90% confidence
Finding: The skill explicitly promotes removing handwritten answers from completed exam papers to recreate blank forms. That normalizes deceptive document alteration and can facilitate academic fraud, falsification, or tampering with records by turning completed materials into apparently original clean copies.
Content: No source excerpt is available for this finding.

Ssd 4

Medium

Category: Not specified by scanner
Confidence: 93% confidence
Finding: The skill presents watermark removal as a standard supported scenario, including a direct example of removing a website watermark. This can enable copyright evasion, provenance stripping, or deceptive redistribution of protected images, making misuse more likely in practice.
Content: No source excerpt is available for this finding.

Vague Triggers

Medium

Category: Not specified by scanner
Confidence: 88% confidence
Finding: The fallback rule for scan-document activates whenever the user expresses a generic optimization intent without clear constraints. Ambiguous fallback behavior increases the risk of unintended execution on unsuitable inputs, especially in an agent system that may infer image-processing intent too broadly and send sensitive content to the external service.
Content: No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.