Agentic SDLC

An Autonomous Multi-Agent Framework for Self-Healing Documentation & Code Optimization

📖 Overview

Agentic SDLC is a next-generation devtool that employs a swarm of autonomous AI agents to accelerate the Software Development Life Cycle. By analyzing your codebase in real-time, it instantly provides professional documentation, rigorous quality audits, and performance optimization patches—all without human intervention.

Designed for modern developers, it features a security-first Session-Key architecture (zero persistence) and a premium dashboard for visualizing AI insights.

🚀 Try the Live Demo

---

📸 Integrated Dashboard

---

✨ Key Features

🤖 Multi-Agent Swarm

Three specialized agents work in parallel using asyncio to deliver results in seconds:

1. Technical Writer Agent: Transforms raw code into beautiful, structured Markdown documentation.
2. Quality Auditor Agent: Performs static analysis to score Code Accuracy, Completeness, and Readability on a 10-point scale.
3. Performance Engineer Agent: Identifies O(n) bottlenecks and proposes optimized, pythonic refactors.

🔒 Privacy & Security

• Session-Only Mode: Your API Key is injected into the browser session RAM and never saved to the server disk.
• Ephemeral Processing: Code is analyzed in memory and discarded immediately after report generation.

⚡ Modern UX

• Real-time Diff Viewer: Compare original code vs. AI-optimized patches side-by-side.
• Glassmorphism Design: sleek Dark/Light mode interface built with Tailwind & CSS variables.
• Interactive Feedback: Visual loading steps showing which agent is currently active.

---

🏗 System Architecture

The project follows a decoupled formatting architecture for maximum scalability:

• Backend (Python/FastAPI):

  • Handles Agent Orchestration via google-genai SDK.
  • Implements robust error handling and fallback logic (Gemini 2.5 Flash -> 2.0 -> 1.5).
  • Exposes a stateless REST API via uvicorn.

• Frontend (Next.js 14):

  • Built with the App Router and Server Components.
  • Responsive layout with custom Inter and JetBrains Mono typography.
  • Seamlessly connects to any deployed backend via Environment Variables.

---

🚀 Quick Start

Prerequisites

• Python 3.9+
• Node.js 18+
• A Google Gemini API Key

1. Clone & Setup Backend

git clone https://github.com/sciro24/Agentic-SDLC.git
cd Agentic-SDLC

# Create Virtual Env
python -m venv venv
source venv/bin/activate

# Install Deps
pip install -r requirements.txt

# Run Server (Port 8000)
./venv/bin/uvicorn src.app:app --reload

2. Setup Frontend

cd web
npm install

# Run Client (Port 3000)
npm run dev

3. Usage

Open http://localhost:3000. You will be prompted to enter your Google API Key securely. Once entered, drag & drop any .py file to start the swarm!

---

👨‍💻 Credits

Architected and Developed by Diego Scirocco. Powered by Google Deepmind Gemini Models.