Enable Gemini multimodal analysis in Claude Code workflows via a Node.js CLI.
npx skills add https://github.com/dev-jelly/jelly-dotclaude --skill jelly-gemini
Please help me install this Agent Skill. Skill: jelly-gemini Source: https://github.com/dev-jelly/jelly-dotclaude/tree/main/skills/jelly-gemini Command: npx skills add https://github.com/dev-jelly/jelly-dotclaude --skill jelly-gemini
This Skill exposes Google's Gemini AI models to Claude Code, enabling multimodal analysis, long-context processing, and fast inference for complex tasks.
Initialize the Gemini CLI or gemini wrapper and run a prompt with the desired model.
High-intent search queries and answers about installing and using this skill.
Gemini-based multimodal analysis lets you process text and images together through Claude Code. Initialize the Gemini CLI wrapper, provide your text prompt with optional images, and select your model (gemini-2.5-flash, gemini-2.5-pro, or gemini-3.0-deep-think) to get streaming results formatted as text, JSON, Markdown, or code.
Yes. Gemini supports long-context processing up to 1,048,576 tokens, making it suited for large document processing, full codebase review, and architecture planning tasks that exceed standard token windows.
The Skill requires the @google/generative-ai dependency and environment-key authentication. Set up your Google API key, initialize the Node.js CLI wrapper, and you're ready to run multimodal prompts with streaming support and error handling built in.
Multimodal analysis covers code reviews, UI/UX analysis, documentation generation, and architecture planning. You can pair code snippets with screenshots, diagrams, or design files to get comprehensive feedback in a single Gemini request.
Model choice depends on your latency and reasoning needs. Gemini-2.5-flash prioritizes speed for routine tasks, gemini-2.5-pro balances performance and depth, and gemini-3.0-deep-think delivers advanced reasoning for complex analysis at higher latency.
Gemini outputs results in text, JSON, Markdown, or code formats. This flexibility lets you stream responses directly into documentation, feed structured JSON into downstream tools, or extract formatted code snippets for immediate use.
brixtonphamProcess audio, image, video, and documents via Gemini's multimodal API.
expectedparrotParse EDSL Results by UUID and generate Markdown, HTML, and CSV analysis packages.
alex-tgkAnalyze local and YouTube videos via the Gemini API for summaries, transcripts, and timestamped insights.
zzafergokAccess Google Gemini models via CLI or API for reasoning and multimodal tasks.
