chatgpt-image-generator
ChatGPT Image Generator
Use this skill when the user wants image generation through the ChatGPT web UI and wants the results saved locally.
This skill coordinates the browser workflow; it does not use the OpenAI API. It relies on the user's own ChatGPT browser session and never asks for or stores account credentials.
Conversation Routing
Before uploading files or sending a prompt, call classifyChatGPTConversationMode(task) or prepareChatGPTConversation(tab, task) when practical. Route before any external side effect.
The router returns one of these actions:
new_temporary: open a clean Temporary Chat for a self-contained, one-turn task that does not need image generation.new_standard: open a clean standard chat for image generation, research, campaigns, batches, or future iteration.reuse_current: keep the current conversation when the task depends on its history/attachments, or when performing one bounded mechanical Retry.blocked: stop when an explicit Temporary Chat request conflicts with required existing context or an unsupported capability.
Current verified ChatGPT behavior does not support image generation in Temporary Chat. Route image requests to new_standard, even when the image task is small. Temporary Chat is useful for small one-off text tasks such as concise rewriting, extraction, classification, or translation. Treat this as a capability boundary that may change; do not infer support without verification.