| Project ID |
BITS-SRIP/62A1C8/2026 |
| Project Title |
Agentic Cyber-Defense: Building an Autonomous AI Interface for Kali Linux |
| Project Description |
As Generative AI evolves, the next frontier is Agentic AI—systems that dont just chat, but actively use tools to solve problems. This project offers a unique opportunity to build an Agentic Cybersecurity Tool from the ground up. The goal is to democratize access to powerful security testing by creating a Conversational Web Interface that sits on top of the industry-standard Kali Linux operating system.
Currently, using security tools requires memorizing complex command-line syntax. In this 2-month internship, you will engineer a solution where a user can simply ask an AI agent to scan the network for open ports or check this API for vulnerabilities, and the system will autonomously execute the underlying Kali Linux tools to perform the task.
The project is strictly engineering-focused. You will work on integrating Large Language Models (LLMs) with a Dockerized environment to create a functional prototype.
Key Engineering Modules:
1. Environment Setup: Deploying and managing a Dockerized Kali Linux container on Ubuntu 24.04 LTS to serve as the toolbox for the AI agent.
2. Backend Development (Python): Writing the orchestration logic that translates natural language prompts from the LLM (using Ollama for offline or APIs like Gemini/OpenAI) into executable Linux shell commands.
3. Frontend Development (Zero-Code/React): Using prompt-to-code methodologies to rapidly build a React-based UI/UX. The focus is on creating intuitive Chat and Dashboard interfaces.
4. Tool Integration: Wiring up specific security tools (e.g., Nmap for network, OWASP ZAP for web apps) to be triggered by the Agent.
Expected Tangible Outcomes:
1. Functional Web Application: A deployed React + Python application where a user can chat with the system to trigger at least 3 distinct security tools within the Docker container.
2. Agentic Workflow Engine: A working backend module capable of taking a user intent (e.g., Analyze this URL) and converting it into a structured command sequence without human intervention.
3. Secure Sandbox Implementation: A configured Docker environment that ensures the AI agent executes commands safely within the containerized Kali instance.
4. Technical Documentation: A comprehensive GitHub repository containing the source code, setup scripts, and a How-to guide for future developers to extend the tool.
This project focuses on building the foundation. You will not be writing policies or doing high-level compliance; you will be writing the code that makes the AI Agent hands-on. |
| Project Discipline |
Computer Science / Software Engineering / AI Machine Learning / Cybersecurity (with strong coding fundamentals) |
| Faculty Name |
Ashutosh Bhatia |
| Department |
Department of Computer Science & Information Systems |
