Murtaza Shabir
Cybersecurity & Applied AI Security
Murtaza ShabirCybersecurity & Applied AI Security.
High-school student & independent researcher exploring penetration testing, wireless security auditing (802.11), and adversarial LLM defenses. Focused on foundational mechanics rather than superficial shortcuts.
Foundational mechanics over
superficial shortcuts.
I am Murtaza Shabir, a Class 11 STEM student and independent security researcher. I believe true technical depth comes from understanding core principles from the ground up, not merely stringing together pre-built tools.
While keeping pace with high-school mathematics, physics, and computer science, I spend my lab hours exploring penetration testing, wireless protocol auditing, and large language model vulnerabilities.
Whether analyzing raw 802.11 4-way handshakes, mitigating bufferbloat with custom SQM queue policies, or testing adversarial jailbreaks against LLM endpoints, the goal is always the same: know the mechanics thoroughly, build custom tooling, and verify every hypothesis.
“True security and systems resilience aren't achieved through surface-level shortcuts. They are engineered through rigorous first-principles understanding—from RFC packet frames to neural tokenizers.”
Class 11 STEM Rigor
Balancing formal high school coursework with advanced systems research. Calculus, mechanics, and discrete algebra establish the mathematical bedrock needed for network queue theories and cryptography.
Hands-on Security & CTFs
Competing in Capture The Flag challenges and attacking isolated lab testbeds. Developing an attacker's sharp intuition to engineer resilient, provable defense systems.
Adversarial LLM Defenses
Auditing transformer APIs against direct/indirect prompt injection, multi-turn jailbreaks, and automated exfiltration vectors. Engineering deterministic semantic guardrails.
RFC & Protocol Analysis
Zero superficial shortcuts. Reading IEEE 802.11 standards, dissecting 4-way handshakes in Wireshark, writing custom Scapy scripts, and tuning OpenWrt CAKE SQM policies.
Domains & Technical Capabilities
An engineering-driven skill matrix spanning offensive wireless auditing, machine learning defense mechanisms, and network subsystems.
Offensive & Defensive Security
Methodological security testing from 802.11 wireless frame dissection to penetration testing attack cycles.
- Wireless protocol auditing (IEEE 802.11 standards)
- Packet dissection & traffic analysis (Wireshark, Scapy)
- Penetration testing methodologies & reconnaissance
- Competitive CTF problem-solving & binary exploit logic
- Kali Linux operational toolchain
AI & LLM Security
Adversarial threat modeling for language model APIs, multimodal pipelines, and autonomous agents.
- Threat modeling against prompt injections & jailbreaks
- Adversarial testing of multimodal pipelines
- Semantic output sanitization & vector guardrails
- Automated API auditing harnesses
Systems & Networks
Low-level system tuning, queue theory, and bufferbloat elimination on bare metal and embedded devices.
- Linux environments (Kali Linux, Debian, Arch toolchain)
- Containerized security testing environments in Docker
- OpenWrt router configurations & network policy
- SQM QoS bufferbloat mitigation (CAKE, fq_codel)
- Low-level Linux kernel network tuning
Scripting & Tooling
Automation-first software engineering to build reproducible security pipelines and telemetry bots.
- Python (Scapy for packet crafting, AsyncIO concurrency)
- Playwright headless browser automation & session rotation
- Bash shell scripting & system automation
- Wireshark dissectors & pcap file analysis
- Git version control & modular testing workflows
Applied Projects & Research Labs
Tangible systems engineered to investigate threat surfaces, test defensive boundaries, and automate critical telemetry workflows.
LLM Adversarial Assessment & Defensive Guardrails
Automated evaluation framework designed to test large language model endpoints against prompt injection, multi-turn jailbreaks, and indirect data exfiltration techniques.
Dynamic evaluation harnesses, semantic vector guardrails, and anomaly validation for untrusted prompt inputs. Implements automated test harnesses using PyTest and isolated container execution.
Automated Wireless Auditing & Network Latency Lab
Scripted security auditing pipeline for 802.11 wireless frame verification combined with Smart Queue Management (SQM) QoS profiling to eliminate bufferbloat under saturated load.
Headless automation workflows for network reconnaissance and handshake verification alongside custom OpenWrt network policies and Linux tc/CAKE queue disciplines.
Headless Recon & Telemetry Engine
Continuous endpoint monitoring and intelligence gathering engine with real-time telemetry dispatch and anomaly notifications.
Playwright-backed headless pipeline with dynamic fingerprint rotation, session management, and automated Telegram Bot API telemetry alerts.
Connect & Collaborate
Open to technical discussions, competitive CTF team-ups, security research opportunities, and academic dialogues.
GitHub
Codebase, repositories, and open-source lab experiments
Instant messaging for fast project discussion & direct chats
Telegram
Fast encrypted messaging channel for inquiries & collaboration
Direct email for academic inquiries & formal communication