ClawHub
Back to skill

Security audit

Vision

Security checks for vulnerabilities and agentic risk

Overview

This skill is a local image-processing helper whose file access and metadata reading are related to its purpose, with some documentation gaps users should notice.

Install only if you are comfortable letting the agent run local ImageMagick commands on image files you choose. Be careful with the info command because photo metadata can reveal camera/device details and timestamps; review outputs before sharing them.

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
  • 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
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (4)

Tp4

High
Category
MCP Tool Poisoning
Confidence
95% confidence
Finding
Most of the code aligns well with the declared image-processing purpose: it resizes, crops, converts formats, optimizes files, and adds text watermarks using ImageMagick. However, the code also provides a distinct `info` command for extracting image metadata and EXIF data, including camera make/model, date taken, exposure, ISO, and focal length. That capability is not represented in the declared description, which focuses only on transformation/optimization tasks. While related to images, metadata extraction is a separate functional capability and should be disclosed. The script also conditionally invokes `exiftool`, which is not mentioned in the description.

Missing User Warnings

Medium
Confidence
90% confidence
Finding
The info command extracts and displays EXIF metadata, including camera make/model and capture date, and can emit it in JSON without any warning, consent check, or redaction option. In an agent context, this can unintentionally disclose sensitive image metadata such as device details, timestamps, and potentially other identifying information if users process personal photos.

Missing User Warnings

Low
Confidence
85% confidence
Finding
The markdown describes resize, convert, optimize, and watermark operations, all of which affect user files, but it does not include any user-facing warning about file modification or output handling. Because this skill operates on local image data and may replace or generate files, users should be explicitly informed of that behavior.

Natural-Language Policy Violations

Low
Confidence
78% confidence
Finding
The quick-start section mixes English and Chinese instructions, which imposes a specific locale presentation without stating that language is optional or user-selectable. Under the language/locale policy, user-facing language constraints or assumptions should be explicit or opt-in.

Static analysis

No suspicious patterns detected.