Secure.
Intelligent.
Autonomous.

Exploring the Future

I’m Manuel Nájera — cybersecurity and AI enthusiast exploring the intersection between intelligent automation and digital defense.

I’m currently researching and experimenting with tools, agents, and protocols like:

  • 🧠 LLM-based AI agents such as Goose and UI-TARS-desktop
  • 🧩 Automation frameworks including n8n, Google A2A, and MCP
  • 🛡️ Ethical hacking practices, while completing my Cisco Ethical Hacking Certificate

My goal? To build autonomous systems that are secure by design.


⚡ Quick Look: More tools I’m exploring

  • CUA: High-speed virtual container framework for AI agents to control full operating systems at near-native performance.
  • Reaper: AI-ready, open-source application security testing framework by Ghost Security.
  • Skyvern: Automates browser-based workflows using LLMs and computer vision.
  • TARS: Experimental framework to automate parts of penetration testing using AI agents.

🧭 What you’ll find here

  • Notes and insights from my journey into AI and cybersecurity
  • Project ideas, experiments, and prototypes in progress
  • Tools I discover and test along the way

Whether you’re a fellow builder, recruiter, or simply curious — I hope you find something valuable here.

➡️ Check out my Projects or recent Blog posts.