패키지, 저장소 및 앱 카테고리 검색

GitHub 저장소galaxyrayyan/MoltGuard-Security

MoltGuard-Security

Gemini 2.5 Flash로 구동되는 AI 에이전트용 모바일 보안 오케스트레이터. 실시간으로 프롬프트 주입 및 데이터 유출 감지

MoltGuard-Security 프로젝트 이미지
스타
1
포크
0
최근 푸시(UTC)
2026. 3. 21.
프로젝트 상태
활성

원본 README

아래는 영문 원문 스냅샷입니다. 최신 내용은 GitHub에서 확인하세요.

🛡️ MoltGuard: AI Security Orchestrator

Flutter Gemini 2.5 Flash License: MIT

If you find this project useful, please consider giving it a ⭐ to support its development!

MoltGuard is a cutting-edge mobile security orchestrator powered by the next-generation Gemini 2.5 Flash. It performs high-speed behavioral auditing on AI interactions to detect prompt injections, malware patterns, and data leaks with unprecedented reasoning capabilities.

⚙️ Quick Setup & API Configuration

To get started with MoltGuard, follow these steps:

  1. Obtain API Key: Get your free API key from Google AI Studio.
  2. Environment Setup: Create a .env file in the root directory.
  3. Add Key: Add your key to the file: GEMINI_API_KEY=your_key_here.
  4. Run: Execute flutter run in your terminal.

🧠 How the Security Analysis Works

MoltGuard acts as a middleware between the user and the LLM.

  1. Input Sanitization: User prompts are intercepted.
  2. Behavioral Auditing: Gemini 2.5 Flash analyzes the intent (detecting Jailbreaks, PII Leaks, or Malicious Code).
  3. Structured Response: The AI returns a precise JSON object containing:
    • overallRiskScore: Numerical value (0-100).
    • vulnerabilities: Detailed list of detected threats.
  4. Actionable Output: Users receive a visual risk level and a downloadable PDF report.

🚀 Key Features (Powered by Gemini 2.5 Flash)

  • Extreme Reasoning Speed: Leveraging the ultra-fast Gemini 2.5 Flash for real-time threat analysis without latency.
  • Lux Loading Radar: A premium security scanning animation for an immersive "Cybersecurity" experience.
  • Smart Risk Classification: Internal logic that auto-calculates threat levels (Critical, High, Medium, Low) ensuring no "Unknown" results.
  • Professional PDF Audits: Generate detailed security reports with QR Code verification, powered by deep AI insights.
  • Modular Clean Architecture: Organized into a professional Service Layer (Models, Services, Screens) for enterprise scalability.

📸 Preview

AI Radar Scanning Threat Analysis Result PDF Export Feature

Security Dashboard Scan History

MoltGuard in action: Real-time auditing and professional reporting powered by Gemini 2.5 Flash.

📂 Project Structure

lib/
├── models/      # Data Structures & Smart Fallback Logic
├── services/    # AI Service (Gemini 2.5), Export (PDF), & Storage
├── screens/     # Premium UI Components
└── main.dart    # Application entry point

🗺️ Future Roadmap

We have ambitious plans for MoltGuard! From on-device ML to enterprise-grade security monitoring. 🚀 View our detailed Roadmap here


🛠️ Built With

  • Flutter - High-performance core cross-platform framework.
  • Gemini 2.5 Flash - Utilizing the next generation of AI for advanced safety auditing.
  • FlutLab - Professional cloud IDE used for rapid development and orchestration.
  • MIT License - Open for community contribution and safety research.

⚠️ Disclaimer

This tool is for educational and research purposes only. Always use AI responsibly and ensure you comply with data privacy regulations.


Note: This project showcases the cutting-edge application of Gemini 2.5 Flash in the field of AI safety and mobile security guardrails. It aims to empower users with a transparent layer of protection in an AI-driven world.

Irayyan GitHub avatar
GitHub 작성자

Irayyan ↗

언어Dart

분류

사용 중인 의존성

의존성 목록 9 개

프로젝트에 맞는 저장소를 선택하려면?

사용 목적에 따라 검색하고 개발 활동을 비교한 뒤 원본 저장소에서 요구 사항을 확인하세요.

  1. 이름, 작성자 또는 설명으로 검색하고 기술 주제로 범위를 좁힙니다.
  2. 스타, 포크, 최근 푸시를 함께 비교하세요. 스타는 관심의 지표이며 품질을 보장하지 않습니다.
  3. 사용하기 전에 GitHub에서 README, 라이선스, Issues를 확인하세요.

출처 및 관리

“A repository is the most basic element of GitHub.”

GitHub Docs — About repositories ↗