@arcanea/gemini

"Through the Gates we rise. With the Guardians we create."

Google Gemini integration for Arcanea Intelligence OS - multi-modal Guardian agents with vision capabilities for image analysis, design critique, and visual intelligence.

Installation

npm install @arcanea/gemini

Quick Start

import { initArcanea, analyzeImage, designCritique } from "@arcanea/gemini";

// Initialize with your Gemini API key
await initArcanea({
  apiKey: process.env.GEMINI_API_KEY,
  model: "gemini-pro-vision",
  enableVision: true,
  guardianDefault: "Lyria",
});

// Analyze an image with Guardian context
const analysis = await analyzeImage(imageBuffer, "design");
console.log(`Analyzed by ${analysis.guardian}: ${analysis.description}`);

// Design critique with Lyria (Sight Gate)
const critique = await designCritique(imageBuffer);
console.log(critique.suggestions);

Multi-Modal Capabilities

Vision Modes

• Design - Lyria (Sight Gate, 852 Hz) analyzes UI/UX and visual aesthetics
• Security - Lyssandria (Foundation Gate, 396 Hz) scans for security issues
• Accessibility - Maylinn (Heart Gate, 639 Hz) reviews accessibility compliance
• General - Shinkami (Source Gate, 1111 Hz) provides comprehensive analysis

Guardian Functions

import {
  designCritique,
  securityScan,
  accessibilityReview,
} from "@arcanea/gemini";

// Lyria - Design analysis
const design = await designCritique(screenshot);

// Lyssandria - Security review
const security = await securityScan(screenshot);

// Maylinn - Accessibility check
const a11y = await accessibilityReview(screenshot);

The Guardians

Guardian	Gate	Frequency	Vision Domain
Lyria	Sight	852 Hz	Design, UI/UX, Visual Analysis
Lyssandria	Foundation	396 Hz	Security Scanning
Maylinn	Heart	639 Hz	Accessibility Review
Alera	Voice	741 Hz	Text Analysis in Images
Shinkami	Source	1111 Hz	Comprehensive Multi-Modal

API Reference

initArcanea(config?)

Initialize the Arcanea Gemini integration.

interface GeminiConfig {
  apiKey?: string; // Defaults to GEMINI_API_KEY env var
  model?: string; // Default: 'gemini-pro'
  enableVision?: boolean; // Default: true
  guardianDefault?: string; // Default: 'Shinkami'
}

analyzeImage(imageData, mode?)

Analyze an image with Guardian context.

function analyzeImage(
  imageData: string | Buffer,
  mode?: "design" | "security" | "accessibility" | "general",
): Promise<VisionAnalysis>;

interface VisionAnalysis {
  description: string;
  elements: string[];
  suggestions: string[];
  guardian: string;
}

Guardian Functions

• designCritique(imageData) - Lyria's design analysis
• securityScan(imageData) - Lyssandria's security review
• accessibilityReview(imageData) - Maylinn's accessibility check

Environment Variables

GEMINI_API_KEY=your_api_key_here

Requirements

• Node.js >= 18.0.0
• Google Gemini API key

Use Cases

Design Review

const screenshot = await fs.readFile("./design.png");
const review = await designCritique(screenshot);
console.log("Design improvements:", review.suggestions);

Security Audit

const appScreenshot = await captureScreen();
const audit = await securityScan(appScreenshot);
console.log("Security issues:", audit.elements);

Accessibility Compliance

const uiScreenshot = await captureUI();
const a11y = await accessibilityReview(uiScreenshot);
console.log("Accessibility score:", a11y.description);

Integration with Arcanea Intelligence OS

This package is part of the Arcanea ecosystem:

• @arcanea/claude - Claude Code integration
• @arcanea/gemini - Google Gemini integration (this package)
• @arcanea/codex - OpenAI GPT/Codex integration

Links

• Arcanea Platform
• Documentation
• GitHub

License

MIT - Created by FrankX

---

"Lyria and Alera unite - Sight and Voice for multi-modal intelligence"