baoyu-danger-gemini-web

Automate Gemini text generation by passing prompts, prompt files, or stdin to a script that manages browser authentication and session state. You receive formatted text outputs directly in your terminal without manual web interaction.

Reference images can be used as input for Gemini vision analysis by passing them through the automated workflow. The process handles image uploads, executes vision-enabled chats, and returns analyzed text content.

Maintain multi-turn Gemini conversation context using persistent session IDs and saved message history. This allows subsequent prompts to reference prior interactions, preserving the dialogue state across multiple automated turns.

You do not need a standard Gemini API key to generate images on demand. The workflow requires browser-authenticated Gemini cookie management to access web features, handling authentication internally without traditional API credentials.

Save generated Gemini images locally by specifying an output image path in your command configuration. The workflow intercepts returned image data from the generation process and writes the files to your designated directory.

Using browser cookies for Gemini automation requires active session management and periodic cookie refreshes to prevent expiration. This approach depends on web interface stability and may lack the guaranteed rate limits of a dedicated API client.