Chatgpt Image Generation
Security checks across static analysis, malware telemetry, and agentic risk
Overview
This skill appears to do what it claims, but users should notice that it installs Playwright, uses a logged-in ChatGPT session, sends prompt files to ChatGPT, and logs prompts locally.
Install only if you are comfortable adding Playwright/Chromium, logging into ChatGPT in an automated browser, and sending the prompt file contents to ChatGPT. Keep prompts non-sensitive and protect the output directory because results.jsonl records the prompt text.
Static analysis
No static analysis findings were reported for this release.
VirusTotal
VirusTotal findings are pending for this skill version.
Risk analysis
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
Installing the skill setup dependencies brings external code onto the machine.
The skill requires installing an npm package and a Chromium browser binary. This is expected for Playwright browser automation, but it depends on external package/browser supply chains.
npm install playwright npx playwright install chromium
Install from a trusted environment, consider pinning Playwright versions, and avoid running the setup with elevated privileges unless necessary.
Generated requests may be associated with the user's ChatGPT account and may consume account usage limits or credits.
The script operates through a ChatGPT browser session and asks the user to sign in when needed. This is purpose-aligned, but actions occur under the user's ChatGPT account.
console.log('⚠️ Not logged in! Please sign in to ChatGPT, then press Enter...');Use an account appropriate for automation, review ChatGPT/OpenAI usage policies, and avoid using credentials or sessions you do not want the automation to access.
Anything in the prompt file is transmitted to ChatGPT through the web UI.
The script sends each prompt from the local prompts file to ChatGPT. This external provider communication is central to the skill, but users should treat prompt contents as shared with ChatGPT.
await page.goto('https://chatgpt.com/', ...);
await textarea.fill(prompt);Do not include secrets, confidential business data, or private personal information in prompts unless you are comfortable sharing that data with ChatGPT.
Prompt contents may remain on disk in the output directory after the image generation run finishes.
The script writes full prompt text into a persistent local results.jsonl log. This is disclosed by the output behavior, but it can retain sensitive prompt content.
logResults(resultsPath, { index: i, prompt, status: 'success', output: outputPath });Choose an appropriate output directory, protect or delete results.jsonl if prompts are sensitive, and avoid storing confidential prompts in shared folders.