I build systems, then I try to break them before someone else does.
Mohit Sharma — full-stack developer with a focus on offensive security. I ship working products end to end and spend the rest of my time figuring out how software fails, so the things I build don't.
What I've shipped
Self-Heal-AI
security · automationAn AI-assisted self-healing agent for servers — watches logs, flags anomalies, blocks malicious IPs automatically, restarts failed services, and alerts over Telegram/Discord. Built to close the gap between "detected" and "fixed."
Vulnerability Scanner
offensive securityA scanner built to identify common exploitable weaknesses in target applications — misconfigurations, exposed endpoints, and known vulnerability signatures — with results structured for direct use in a report.
Exbow-Offensive
offensive security researchA recon-to-report automation pipeline for ethical hacking workflows — reconnaissance, analysis, and exploitation testing against consenting, authorized targets, with structured reporting for write-ups and CTFs.
ExamCraft
AI toolA marks-aware answer sheet generator for university students — reads a question paper, scales responses to the marks allotted, walks through maths step by step, and exports a properly formatted A4 PDF. Single-file, no backend required.
Quiz-Master Pro
real-time · multiplayerA multiplayer quiz game with a live leaderboard, AI-generated explanations for wrong answers, and gamified badges to keep people coming back.
How I got in
Vulnerability Scanner — findings writeup
CTF / disclosure styleFull breakdown of how the target's weak point was found, how it was confirmed and exploited under authorization, and what fixed it — written the way a real disclosure report reads, not a diary.
Read the writeup →Why me?
What I work with
LANGUAGES
- Python
- C
- JavaScript
WEB & TOOLS
- Tailwind CSS · React
- Supabase
- Git / GitHub
SECURITY
- Recon & enumeration automation
- Vulnerability scanning & reporting
- Log-based anomaly detection
- CTF-style exploitation practice
AI / ML APPLIED
- LLM-integrated tooling (GPT-4o family)
- OCR & document pipelines
- Practical, product-shaped AI features