ClawHub
Back to skill

Security audit

GPT Image Edit — Pro Pack on RunComfy

Security checks for vulnerabilities and agentic risk

Overview

This skill is a clearly disclosed wrapper for RunComfy image editing, with no hidden scripts or unrelated behavior found.

Install this only if you are comfortable sending your edit prompt and publicly fetchable image URLs to RunComfy for processing. Use explicit invocations and avoid submitting private images or sensitive embedded text unless that sharing is acceptable.

SkillSpector

Vulnerability Patterns
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • 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 (1)

Vague Triggers

Medium
Confidence
94% confidence
Finding
The skill declares it should trigger on both several named phrases and also on "any explicit ask to edit with this model," which is an overly broad activation rule. Broad triggers can cause the skill to activate in contexts the user did not clearly intend, leading to unintended execution of a third-party CLI, transmission of user-provided image URLs and prompts to a remote service, and possible misuse of local credentials such as RUNCOMFY_TOKEN.

Static analysis

No suspicious patterns detected.