ClawHub
Back to skill

Security audit

零依赖图片诊断与并排对照（无 Pillow）

Security checks for vulnerabilities and agentic risk

Overview

This skill is a disclosed local PNG inspection and transformation helper with file reads and output writes that match its stated image-diagnostics purpose.

Install this if you want a local, no-dependency PNG inspection tool. Be aware that output commands can overwrite the path you provide with -o, and the tool is documented primarily in Chinese and supports PNG only, not JPEG without prior conversion.

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
  • 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 (4)

Tp4

High
Category
MCP Tool Poisoning
Confidence
91% confidence
Finding

The declared description presents the skill as both an image processing and image diagnostics tool, emphasizing analytical features for clarity, color, banding, and comparison. The supplied code only covers a subset: PNG decode/encode, RGB conversion, bilinear resize, crop, non-white bounding box detection, fit-to-width, and vertical stacking. These are consistent with part of the description, but several prominent declared capabilities are absent, especially the diagnostic/statistical functions and simulated dark-region lifting. There is no evidence of undeclared sensitive behavior or inconsistent resource access; the mismatch is that the description significantly overstates what the code actually does.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

该 markdown 文件在“何时使用”描述中把“帮我对比这两张图”“颜色不对”等常见日常说法作为触发示例，但未限定必须是图片诊断场景，也没有给出不应触发的反例或更明确的边界。这类表述容易与一般性的比较、配色讨论或截图处理请求重叠，造成技能被过度匹配。

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
84% confidence
Finding

The module docstring and inline documentation are written entirely in Chinese, which can amount to an implicit language constraint for users or maintainers if the skill is expected to operate in a broader environment. There is no indication that this locale choice is optional, user-selected, or justified as region-specific.

Content

No source excerpt is available for this finding.

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
78% confidence
Finding

The function writes PNG data to an arbitrary filesystem path using open(path, "wb").write(png), but there is no confirmation prompt, logging, or inline warning describing that the skill modifies local files. For a code file, filesystem writes are in scope for missing-warning review when no visible disclosure is present.

Content

No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.