Skip to content
View paultheal1en's full-sized avatar
2 followers · 10 following

Achievements

Achievement: Quickdraw

Achievements

Achievement: Quickdraw

Block or report paultheal1en

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
paultheal1en/README.md
Van Truong — Security, AI and Autonomous Systems

Security builder exploring what happens when AI gets tools, memory, and a permission model.
Information Security @ UIT · building useful autonomy without accidentally inventing Skynet-as-a-Service.


Animated system status console for Security, AI and Autonomous Systems

$ about_me

I'm Van Truong — a security builder who likes giving AI agents tools, constraints, and a reason to distrust their first answer.

I started with offensive security: breaking web applications, studying how systems fail, and automating the repetitive parts. Now I explore the other side of the problem — building autonomous systems that can investigate, reason, and act without becoming the next security incident.

What interests me most is not how intelligent an agent sounds, but whether it can show its evidence, stay inside scope, and stop when it is uncertain.

$ current_rabbit_holes

  • autonomous security workflows that do more than chat
  • tool use, memory, permissions, and agent guardrails
  • turning offensive knowledge into safer automation
  • making AI say “I don't know” before production says “incident”

$ operating_manual

agent = Agent(
    tools="useful",
    permissions="minimal",
    evidence="required",
    human_override=True,
    sudo=False,
)

Working rules: evidence before confidence · autonomy should be reversible · humans keep the final veto · if the YAML works on the first try, inspect it again.

$ tools_i_reach_for

Python · TypeScript · Linux · Docker · LangGraph · MCP · Web Security · Security Automation

currently building things, breaking assumptions, and keeping autonomous systems on speaking terms with reality.

Popular repositories Loading

  1. devign devign Public

    Reimplement partial implementation of Devign, a graph neural network-based model for identifying vulnerabilities in C code. Includes tools for data preparation, Joern integration, and model trainin…

    Python 1

  2. set-up-cryptopp set-up-cryptopp Public

    Makefile

  3. Cryptography_coursework Cryptography_coursework Public

    Reimplementation of Xu et al.'s searchable and privacy-preserving data sharing scheme for cloud-assisted e-healthcare systems. Includes C implementation, Python integration.

    C++

  4. Web-App-Sec Web-App-Sec Public

    A web-based platform for cooking enthusiasts featuring recipe search, AI-powered recipe creation, calorie calculation using Google Studio AI APIs, and user-friendly tools for saving and exploring r…

    JavaScript

  5. Deep-Learning-based-Vulnerability-Detection-Are-We-There-Yet---Reimplement Deep-Learning-based-Vulnerability-Detection-Are-We-There-Yet---Reimplement Public

    Reimplement empirical study on deep learning-based vulnerability detection techniques using real-world datasets (Devign and Chrome+Debian). Includes tools for parsing, slicing, and analyzing C code…

    C 1

  6. Deploy-Lavarel-NT213 Deploy-Lavarel-NT213 Public

    Forked from PhanQDat/Deploy-Lavarel-NT213

    A website for understanding, exploiting, and preventing common vulnerabilities in Laravel applications. Includes examples, prevention techniques, and best practices for secure Laravel development.

    PHP