Mastodawn

Step 1: Install an AI agent on a Raspberry Pi 5
Step 2: Hope it does not become a security incident
Step 3: Secure it properly

How I installed and hardened OpenClaw (Moltbot / Clawdbot):

https://blog.shellntel.com/p/installing-openclaw-moltbot-clawdbot-securely-on-a-raspberry-pi5

#AI #Cybersecurity #RaspberryPi

Installing OpenClaw (Moltbot/Clawdbot) Securely on a Raspberry Pi 5

Write Code While You Sleep

#_shellntel Cybersecurity Blog

Ryan Zagrodnik Jan 2

Someone will remember you not for what you hacked, but how you shared it.

Ryan Zagrodnik Dec 8

Ryan Zagrodnik Aug 28, 2025

Check out my latest blog post https://blog.shellntel.com/p/def-con-33-and-meshtastic-on-the-lilygo-t-deck-plus

DEF CON 33 and Meshtastic on the LILYGO T-Deck Plus

Real-world testing with 2,000+ nodes at the world's largest hacker conference.

#_shellntel Blog

Ryan Zagrodnik Jul 22, 2025

Just claimed my quadrant for #DEFCON33’s World War Drive!
I’ll be wardriving the Madison isthmus – campus, Capitol, State St, and everything in between.
APs beware. Logs incoming. 📡🚗💥
Join the hunt: https://wigle.net/contest/DC33
#WiGLE #WWWD #cybersecurity #wardriving

Ryan Zagrodnik Jul 16, 2025

Used Claude's context resets + its own advice to bypass safeguards via email, leading to RCE through chained trust in Gmail, Claude Desktop, and a shell agent. Wild LLM composability flaw.
https://www.pynt.io/blog/llm-security-blogs/code-execution-through-email-how-i-used-claude-mcp-to-hack-itself

Code Execution Through Email: How I Used Claude to Hack

This is the story of how I used a Gmail message to trigger code execution through Claude Desktop, and how Claude itself (!) helped me plan the attack.

Ryan Zagrodnik Jul 11, 2025

Check out my latest blog post https://blog.shellntel.com/p/meshtastic-2-months-later