Skip to content

Writing

Long-form write-ups of my projects and research.

4 posts

  1. Ghost in the Template: Watching a Chat-Template Backdoor Fire, Live

    You can backdoor a local LLM without touching a single weight.

    • AI Supply Chain
    • LLM Security
    • GGUF
    • Jinja2
    • Adversarial ML
  2. AutoRedTeam-LLM: Building an Automated Security Benchmark for Open-Source LLMs

    Open-source LLMs are deployed in production without systematic safety evaluation.

    • Cybersecurity
    • AI/ML
    • Red Teaming
    • Adversarial ML
    • Python
  3. Agentic Network Forensics: Automated PCAP Analysis at Scale

    Modern network forensic investigations face a scaling problem

    • Cybersecurity
    • Network Forensics
    • AI/ML
    • CTF
    • Python
  4. SMS Phishing Detection with Multi-Agent AI: Building a Personal Anti-Scam Analyst

    What if the judgment didn't have to be human?

    • Cybersecurity
    • AI/ML
    • Multi-Agent
    • Phishing
    • Python
Back to the game