ClawHub
Back to skill

Security audit

ImageTranslate.AI

Security checks for vulnerabilities and agentic risk

Overview

This skill is for translating images through ImageTranslate.AI, but it under-labels a billable image-generation action as safe read-only behavior.

Review before installing. Only use this skill when you explicitly want an image translated through ImageTranslate.AI, and confirm the image URL, output settings, and possible cost before running translate_image.

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 (3)

Description-Behavior Mismatch

Medium
Confidence
90% confidence
Finding
The manifest and description frame the skill as suitable for 'searching and reading data,' but the only documented action performs active content transformation and produces a new rendered image. This mismatch can cause an agent or user to treat the skill as read-only and invoke it without the higher scrutiny normally applied to non-read operations, increasing the chance of unintended processing, charges, or data handling.

Intent-Code Divergence

Medium
Confidence
94% confidence
Finding
The safety section states that untagged actions are reads, yet `translate_image` is untagged despite being a processing/generation action that creates a translated PNG. This unsafe classification can bypass confirmation or policy checks built around write/destructive tags and may lead agents to execute billable or privacy-sensitive transformations automatically.

Vague Triggers

Medium
Confidence
83% confidence
Finding
The instruction to use this skill for ANY ImageTranslate.AI request is overly broad and can cause the skill to be selected whenever the product is merely mentioned, regardless of whether the request actually requires connector execution. In context, that broad trigger is more dangerous because the skill performs active image translation rather than passive lookup, so accidental invocation can expose data to a third-party service and incur costs.

Static analysis

No suspicious patterns detected.