Claude Skill MartBrowse skillsWhat is a Skill?
 
← Back to list

Gemini Web Image & Text Generator (danger)

A CLI skill that generates text and images through a reverse-engineered Gemini Web API, with vision input and multi-turn sessions.

Image & VideoAdvanced★ 24,749⑂ 2,764AI score 6/10Last updated: Jul 4, 2026

What it does

  • Drives Gemini's internal web API from the command line for text generation and image generation.
  • Accepts reference images via --reference for vision tasks such as describing a screenshot or creating variations.
  • Persists multi-turn conversations with --sessionId, and offers --json output so other skills can consume results.
  • Lets you pick a model (gemini-3-pro / flash / flash-thinking / 3.1-pro-preview) and configure proxy, data dir and defaults through EXTEND.md or env vars.
  • Requires an explicit consent step on first use (because the API is reverse-engineered) and caches Google cookies after a browser login.

Who it's for

  • Developers who want to wire Gemini into automation without paying for API keys.
  • Anyone needing an image-generation backend for other skills (thumbnails, docs, slides).
  • Intermediate/advanced users comfortable with bun/npx, env vars and browser profiles.
  • ⚠️ Only use it if you accept the terms-of-service and account-suspension risk.

Examples

  1. Blog thumbnail — prompt "a cute cat illustration" with --image thumb.png to write the PNG directly.
  2. Screenshot review — --reference screenshot.png --prompt "point out UI problems" for vision-based feedback.
  3. Iterative planning — call repeatedly with the same --sessionId project-a to keep context while refining a plan.

· · · Install guide · · ·

Install in the Claude app (no terminal)
  1. Download the ZIP with the button below.
  2. In Claude, open Settings → Capabilities and turn on 'Code execution and file creation'. (one time)
  3. Go to Customize → Skills → + → 'Upload a skill' and upload the ZIP.
↓ Download ZIP
Install in Claude Code

Let Claude do it — paste this into Claude Code

Install the skill I found on Claude Skill Mart.
Copy the skills/baoyu-danger-gemini-web folder from the GitHub repo JimLiu/baoyu-skills into my ~/.claude/skills/baoyu-danger-gemini-web/.
When it's done, tell me in one line what this skill can do.

Install with a command instead

git clone https://github.com/JimLiu/baoyu-skills.git && mkdir -p ~/.claude/skills && cp -r baoyu-skills/skills/baoyu-danger-gemini-web ~/.claude/skills/

⚠ This is a third-party skill. Check the source repository before installing.

  1. Prerequisites: install bun (or rely on npx -y bun), have Chrome/Chromium/Edge available, plus a Google account.
  2. Clone the repo: git clone https://github.com/JimLiu/baoyu-skills.git
  3. Copy the skill: mkdir -p ~/.claude/skills && cp -r baoyu-skills/skills/baoyu-danger-gemini-web ~/.claude/skills/
  4. Restart Claude Code, then ask something like "generate an image with Gemini" to trigger the skill.
  5. Accept the disclaimer: on first run you'll be asked to consent to reverse-engineered API usage; read it, then accept to create the consent file.
  6. Log in: a browser window opens for Google auth and cookies are cached. Use --profile-dir if you prefer a dedicated Chrome profile instead of reusing your personal session.
  7. Smoke test: run bun ~/.claude/skills/baoyu-danger-gemini-web/scripts/main.ts "Hello" --json.
View source on GitHub ↗License: MIT