baoyu-danger-gemini-web

Generate text and images via a reverse-engineered Gemini Web API.

2|Updated Mar 23, 2026
One-click install
npx skills add https://github.com/neo-unicorn/neo-skills --skill baoyu-danger-gemini-web-neo-unicorn
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: baoyu-danger-gemini-web
Source: https://github.com/neo-unicorn/neo-skills/tree/main/baoyu-danger-gemini-web
Command: npx skills add https://github.com/neo-unicorn/neo-skills --skill baoyu-danger-gemini-web-neo-unicorn

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

Provides a command-line client and library to generate text and images by interfacing with the Gemini Web endpoints when official APIs are unavailable, enabling multi-turn chats, reference-image vision input, and session persistence without relying on hosted model endpoints.

Core Features & Use Cases

  • Text generation via reverse-engineered Gemini Web API with selectable models and JSON output.
  • Image generation with support for reference images, saving generated images, and handling full-size downloads.
  • Multi-turn conversation support with session files, cookie-based authentication, auto-refreshing cookies, and optional Chrome CDP login flow for retrieving browser cookies.
  • Use Case: run deterministic CLI prompts to produce article content and accompanying images, or use vision prompts to generate image variations.

Quick Start

Run the skill to generate text and an image from a prompt and save the image to generated.png.

Frequently Asked Questions about baoyu-danger-gemini-web

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate text and images via Gemini Web from the command line?

You can generate text and images via Gemini Web by running a reverse-engineered CLI client that handles prompt-driven creative inputs, multi-turn sessions, and file uploads for vision inputs, saving outputs directly to local files.

How does cookie-based authentication work for reverse-engineered Gemini Web APIs?

Cookie-based authentication for the Gemini Web API involves extracting browser cookies, auto-refreshing them during sessions, or optionally using a Chrome CDP login flow to retrieve valid cookies for uninterrupted CLI access.

Can I use reference images for vision input with a Gemini Web reverse-engineered client?

Yes, the Gemini Web reverse-engineered client supports reference images for vision input by handling file uploads, allowing you to generate image variations or prompt-driven text descriptions based on the provided visual data.

What is the best way to maintain multi-turn conversation sessions when using Gemini Web endpoints?

The best way to maintain multi-turn conversation sessions with Gemini Web endpoints is utilizing session files for persistence, ensuring continuous context across multiple prompts while auto-refreshing cookies for authentication.

Does the Gemini Web reverse-engineered API client support model selection and JSON output?

Yes, the reverse-engineered Gemini Web API client supports selectable models for text generation and can output JSON, satisfying technical needs for deterministic CLI workflows and structured data parsing.

What are the limitations of using a reverse-engineered Gemini Web API for text generation?

Limitations of using a reverse-engineered Gemini Web API include dependency on cookie-based authentication, potential instability from endpoint changes, and lack of official support compared to hosted model endpoints.