What I Built

NEXUS — Elite CyberSec AI Agent is an advanced, AI-powered cybersecurity assistant that automates the most tedious parts of a security professional's workflow — threat analysis, penetration testing checklists, log parsing, and incident response report generation.

The idea behind NEXUS is simple: security professionals spend far too many hours chained to their desks, manually sifting through logs, crafting reports, and cross-referencing checklists. NEXUS handles all of that in seconds, freeing up analysts and pentesters to step away from their screens and focus on what really matters — physical security audits, attending conferences, red-teaming in the real world, and yes, touching grass.

Who is it for?

Penetration testers who want instant, checklist-driven tactical guidance

SOC analysts drowning in log data and report writing

Security students looking for a hands-on AI-powered learning companion

Anyone in cybersecurity who wants to reclaim their time from repetitive screen work

Code

SumitNayak1337 / streamlit_app

🛡️ NEXUS — Elite CyberSec AI Agent

this also live at here:https://nexus1337.streamlit.app/

Built by Sumit Nayak, NEXUS is an advanced AI-powered cybersecurity assistant designed to streamline threat analysis, penetration testing workflows, and incident response reports.

📺 Project Demo

Experience NEXUS in action! The following video demonstrates how the AI analyzes logs, screenshots, and generates professional security reports in seconds.

https://github.com/SumitNayak1337/streamlit_app/raw/main/app_video.mp4

✨ Key Features

🧠 Custom Knowledge Base: NEXUS is trained on Sumit's proprietary cybersecurity checklists and portfolio.
🖼️ Screenshot Intelligence: Upload images of findings, recon data, or vulnerability scans—NEXUS will parse and analyze them with high precision.
📄 Auto-Generated DOCX Reports: Select from Offensive (Pentest) or Defensive (SOC) templates to generate high-quality compliance-ready reports.
🔍 Real-time Tactical Guidance: Get exact commands, payloads, and remediation steps mapped to Sumit’s expert checklists.
⚡ Lightning Fast Analysis: Powered by the latest Gemini Pro Vision (genai) SDK…

View on GitHub

How I Built It

NEXUS is built entirely on open-source and open-weight AI infrastructure:

Google Gemini 2.5 Flash (via the google-genai SDK) — powers the core intelligence for chat, image analysis, and report generation. Gemini's API is free-tier accessible, making it an open and accessible choice for the community.

Streamlit — the open-source Python framework that serves as the app's frontend and backend, deployed freely on Streamlit Community Cloud.

Python-docx — open-source library for generating professional DOCX pentest/SOC reports on the fly.

PIL (Pillow) — open-source image processing for screenshot intelligence (users upload recon screenshots and NEXUS parses them).

Custom Knowledge Base — a hand-curated cybersecurity checklist and portfolio (nexus_knowledge.txt) that grounds the AI's responses in real-world, expert-level pentesting methodologies.

Architecture:

User provides their Gemini API key (free to obtain from Google AI Studio)

NEXUS loads a custom system prompt enriched with the proprietary knowledge base

The chat session is initialized with google.genai and supports multimodal input (text + images)

Modular Python modules handle specific workflows: threat_intel, analysis, log_parser, project_manager, and automation

Reports are generated as downloadable .docx files using Offensive (Pentest) or Defensive (SOC) templates

Why Does Open Innovation Matter?

Open innovation was essential for building NEXUS:

Accessibility — By using Gemini's free-tier API and Streamlit's free community hosting, NEXUS is completely free to deploy and use. A closed, paid-only API would have put this tool out of reach for students and independent researchers who need it most.

Transparency — Cybersecurity demands trust. With an open-source codebase, every prompt, every knowledge base entry, and every report template is auditable. Users can verify that NEXUS isn't exfiltrating data or injecting malicious payloads — something impossible with a black-box closed API.

Community-driven improvement — The modular architecture (modules/threat_intel.py, modules/analysis.py, etc.) is designed for community contributions. Anyone can add new analysis modules, expand the knowledge base, or fork the project for their own security workflow.

Customizability — Unlike a closed platform, users can modify nexus_knowledge.txt to inject their own checklists, company-specific policies, or compliance frameworks (PCI-DSS, HIPAA, NIST). This level of customization is only possible because the entire stack is open.

No vendor lock-in — The google-genai SDK is open-source, Streamlit is open-source, and the knowledge base is plain text. If Google changes their API terms tomorrow, migrating to another open-weight model (Llama, Mistral, etc.) is straightforward.

DEV Community

“NEXUS – AI‑Powered Cybersecurity Assistant that Lets You Step Away from the Screen and Touch Grass”