AshlynxCyber Logo
>_CYBERSECURITY PROFESSIONAL // RED TEAMING // AI SECURITY // RESEARCH

Red Teaming, AI Security & Security Research

I’m Hala Naaz, a cybersecurity professional and builder focused on Red Teaming, AI Security, and Security Research. I work on offensive security, web security, AI attack surfaces, security labs, CTFs, and practical security tooling.

HALA@ASHLYNX: ~
hala@ashlynxcyber:~$
$ EXPERIENCE_WITH
Burp Suite
Nmap
Kali Linux
Wireshark
Python
Linux
AWS
Docker
FastAPI
SQL
Burp Suite
Nmap
Kali Linux
Wireshark
Python
Linux
AWS
Docker
FastAPI
SQL
Burp Suite
Nmap
Kali Linux
Wireshark
Python
Linux
AWS
Docker
FastAPI
SQL
$ CAPABILITIES_MATRIX

Security Capabilities & Technical Stack

identity.sh

$ OPERATOR_IDENTITY

Hala Naaz

Cybersecurity | Red Teaming • AI Security • Research

$ LOCATION

Bhubaneswar, Odisha, India

$ AVAILABILITY

Open to Cybersecurity / Red Teaming / AI Security Opportunities

$ CORE_FOCUS

Red Teaming, AI Security & Security Research

$ CURRENT_PRACTICE

Web Security • CTFs • Security Labs • AI Security

hala@ashlynx:~#

Red Teaming & Offensive Security

Offensive Security & Adversarial Testing

Web Application SecurityVulnerability AssessmentReconnaissanceBurp SuiteNmapKali LinuxMetasploitCTF Practice

AI Security

Security of AI Systems & Applications

AI Security ResearchLLM SecurityPrompt InjectionAI Red TeamingRAG SecurityAI AgentsMCP SecurityAI Application Security

Security Research

Research, Analysis & Security Experimentation

Vulnerability ResearchSecurity ResearchTechnical AnalysisCTF ResearchSecurity LabsThreat AnalysisSecurity Write-Ups

Defensive & Cloud Security

Supporting Security Engineering Skills

Threat DetectionSecurity MonitoringSIEMLinuxAWSDockerWiresharkSecurity Engineering
Independent AI Security Research

AUTHORIZED ≠ APPROPRIATE

AI Agent Intent Security, Delegated Authority & Contextual Least Privilege

Why traditional static authorization collapses in autonomous agent systems. Exploring the vulnerability chasm between tools an agent has technical permission to run versus actions aligned with authentic user intent and safety policy.

Excessive AgencyOWASP LLM08
Confused DeputyIndirect Injection
Intent GatesSemantic Verification
AshlynxCyber Research Labs • Active ProgramRead Full Research
Evaluated Project • Score: 8/10

Unit42-FirmwareX

IoT Firmware Vulnerability Analysis Engine

8 / 10

Evaluated against PRD v1.0 specifications via independent dynamic analysis audit. Real QEMU multi-arch user emulation, angr 10.0.5 symbolic execution, and Z3 constraint solving in fail-closed Linux namespaces (unshare -r -n).

Concrete SMT Payload:b'PWND\x00\x00\x00\x00'
Dynamic Registers:33 x86_64 / 17 ARM32
Zero Host Egress:ENETUNREACH Verified
Audit Date: October 29, 2026View Audit Scorecard
>_ $ ORIGIN_STORY // WHOAMI

From Learning to Building & Researching Security

I’m Hala Naaz, a cybersecurity professional building my technical path through hands-on security projects, Red Teaming practice, AI Security research, CTFs, labs, hackathons, and continuous experimentation.

My work focuses on understanding how systems can be attacked, how AI systems can be abused, and how security can be engineered into practical tools and workflows.

Red Teaming & CTFs AI Security Research HomeLabs & 2 Hackathons
$ FEATURED_SECURITY_SYSTEMS

Deployed Systems & Research Builds

Selected offensive AI tools, firmware vulnerability engines, enterprise CTI platforms, and autonomous simulation frameworks.

Explore All 13 Projects

Falcon Strike-AI

> Autonomous Agentic Breach & Attack Simulation (BAS) & CTEM Engine

CrowdStrike-themed adversary emulation platform that constructs directed property attack graphs, synthesizes constrained ATT&CK campaigns via agentic AI, safely executes non-destructive TTPs, and measures SOC MTTD.

Python 3.11+FastAPINext.js 14 App RouterTypeScriptNetworkXDockerGoogle Gemini 3.8/FlashOpenTelemetryReportLabPytest (157 Tests)Vitest (20 Tests)
Submitted2026

Unit42-FirmwareX

> Automated IoT Firmware Attack Surface & Vulnerability Analysis Engine

Automated multi-architecture IoT firmware analysis engine executing binaries via QEMU user-mode and angr symbolic execution in isolated, fail-closed Linux namespaces.

Python 3.12/3.14QEMU User Emulatorsangr 10.0.5Z3 SolverLinux Namespaces (WSL2)Pytest (206 Tests)
Evaluated: 8/102026

ThreatLens CTI Platform

> Enterprise Cyber Threat Intelligence & SOC Platform

Enterprise-grade CTI and SOC platform designed to aggregate, correlate, enrich, score, and operationalize indicators of compromise (IOCs) across security workflows.

FastAPINext.js 16PostgreSQL 16Elasticsearch 8.xPythonTypeScriptDockerCI/CD Pipelines

AgentBreaker

> AI Agent Prompt Injection & Red Teaming Evaluation Framework

An automated, modular offensive security harness for adversarial fuzzing, prompt injection evaluation, tool hijacking analysis, and OWASP LLM01 compliance reporting across local and cloud language models.

Python 3.11+vLLMOllamaPydantic v2DockerGarak ConnectorsPyRIT Modules
Under Development2026

ScamShield AI

> AI Security & Explainable Scam Intelligence Platform

AI-focused scam detection and explainable security intelligence platform built for First Commit — Bharat Builds Tour 2026 (WeMakeDevs × AWS) to analyze suspicious communications.

PythonNext.jsAWS LambdaAmazon API GatewayGoogle GeminiTailwind CSS
Completed2026

Ransomware Detection & Response System (RDRS)

> Local-First Behavioral Anomaly Defense & Automated Quarantine

Endpoint security daemon monitoring filesystem telemetry, computing Shannon entropy spikes, detecting sliding-window write bursts, and isolating suspicious process trees.

PythonFastAPISQLiteDockerWatchdog APITypeScriptNext.js Dashboard
Submitted2026
OPEN TO OPPORTUNITIES

Let's Connect

I'm open to opportunities and technical conversations across Red Teaming, AI Security, Security Research, and Cybersecurity.