🪞 GPT Image 2 — Image Generation via Your ChatGPT Subscription

• Outcome: Pass
• Latest audit: May 29, 2026 at 8:42 PM UTC
• Version:
  1.0.5

Overview

This skill is a disclosed local wrapper for Codex image generation, with a sensitive but purpose-aligned step that reads newly created Codex session logs to recover the generated image.

Install only if you are comfortable with this workflow storing prompts and image-generation artifacts in local Codex session logs under ~/.codex/sessions and then reading the newly created rollout files to extract the image. Avoid using it for highly sensitive prompts or reference images unless that local retention is acceptable.

SkillSpector

By NVIDIA

Vulnerability Patterns

• Data Exfiltration: External Transmission, Env Variable Harvesting, File System Enumeration
• Prompt Injection: Instruction Override, Hidden Instructions, Exfiltration Commands
• Privilege Escalation: Excessive Permissions, Sudo/Root Execution, Credential Access
• Supply Chain: Unpinned Dependencies, External Script Fetching, Obfuscated Code
• Excessive Agency: Unrestricted Tool Access, Autonomous Decision Making, Scope Creep

Findings (1)

Missing User Warnings

Medium

• Confidence: 92% confidence
• Finding: The script deliberately disables ephemeral execution and then scans ~/.codex/sessions to recover generated image data from persisted rollout files. That means prompts, attached image references, and model/session artifacts are stored on disk and later parsed without any explicit user consent or prominent privacy warning, increasing the chance of unintended local disclosure to other processes, backups, or shared accounts.

VirusTotal

66/66 vendors flagged this skill as clean.