Agentic AI Studio: Multi-Agent Software & Security Architect

Agentic AI Studio is an interactive, browser-based multi-agent engineering platform that demonstrates autonomous software architecture design, optimized code generation, and real-time security auditing.

Unlike traditional single-turn LLM assistants that generate isolated code, Agentic AI Studio establishes an autonomous multi-agent debate and consensus pipeline to ensure architectural soundness, algorithmic efficiency ($O(1)/O(N)$), and strict adherence to OWASP Top 10 security standards.

---

🏗️ Multi-Agent Workflow

The pipeline consists of three specialized AI agents working collaboratively:

graph LR
    User["👤 User Prompt"] --> Architect["📐 Architect AI"]
    Architect -- "System Contracts & Graph" --> Coder["💻 Coder AI"]
    Coder -- "Draft Code & Complexity" --> Security["🛡️ Security Audit AI"]
    Security -- "Vulnerabilities Detected" --> Debate{"💬 Revision Loop"}
    Debate -- "Auto-Remediation" --> Coder
    Security -- "100% Compliance" --> Consensus["🏆 Consensus Reached"]

Loading

1. 📐 Architect AI: Decomposes system requirements, establishes API boundaries, designs data flows, and renders interactive Mermaid.js architecture diagrams in real time.
2. 💻 Coder AI: Translates architectural contracts into modular, well-typed code with explicit time and space complexity evaluations.
3. 🛡️ Security & Quality Audit AI: Inspects the Abstract Syntax Tree (AST) and implementation for critical security vulnerabilities (SQL Injection, XSS, race conditions, memory leaks, ReDoS) and forces autonomous remediation until 100% consensus is achieved.

---

✨ Key Features

• Dark Glassmorphism UI: Premium visual aesthetics using modern CSS3 (backdrop-filter: blur(16px)), cyberpunk accents, and a responsive three-column grid layout.
• Interactive Workflow Topology: Visual node network that tracks active agent states, message propagation, and consensus transitions in real time.
• Streaming Reasoning Terminal: Live typewriter console logging internal agent thoughts, security score calculations, and patch notes.
• Dynamic Architecture Rendering: Instant visual system graphs for microservices, Redis rate limiters, guardrail interceptors, and data pipelines.
• Dual Execution Modes:
  • ⚡ Instant Demo Simulation Mode: Pre-computed multi-stage reasoning traces and intelligent offline heuristic evaluation with zero setup overhead.
  • 🟢 Live Google Gemini API Mode: Supports user-supplied Google Gemini (gemini-2.5-flash) API keys for live, real-time LLM multi-agent debates.

---

🛠️ Tech Stack

• Core: HTML5, Vanilla JavaScript (ES6 Modules)
• Styling: Modern CSS3 (CSS Grid, Flexbox, Glassmorphism, Custom Keyframe Animations)
• Visualizations: Mermaid.js for architecture diagrams, Prism.js for syntax highlighting

---

📦 Built-In Engineering Scenarios

Scenario	Domain	Complexity	Key Technical Highlights
High-Throughput Payment Gateway	FinTech / Microservices	Advanced	Sliding-window rate limiting, atomic SETNX Redis locking, Zod schema validation
GenAI LLM Proxy Gateway	AI Security / Guardrails	Intermediate	PII scrubbing, normalization (NFKD) against jailbreak/leetspeak attacks, ReDoS prevention
Distributed LRU Cache	Core Systems / Algorithms	Advanced	$O(1)$ Hash Map + Doubly Linked List, background non-blocking TTL sweeper (.unref())
Buggy Login Auto-Patch	AppSec / Remediation	Critical	Automated refactoring of SQLi/XSS vulnerable legacy code to parameterized queries & Bcrypt

---

🚀 Quick Start & Installation

Because Agentic AI Studio is built as a lightweight, zero-dependency vanilla web application, no package managers or build tools are required.

Local Execution

1. Clone the repository:

   git clone https://github.com/soham-1801/agentic-ai-studio.git
   cd agentic-ai-studio

2. Open index.html directly in any modern web browser or start a local static server:

   python -m http.server 8123

3. Navigate to http://localhost:8123 to launch the platform.

Hosting on GitHub Pages

1. Push the repository to GitHub.
2. Navigate to Settings ➔ Pages.
3. Select main branch under Build and deployment and click Save.
4. Your live environment will be accessible at https://<username>.github.io/agentic-ai-studio/.

---

📜 License

This project is open-source and licensed under the MIT License. Feel free to use, modify, and distribute it across personal or commercial projects.