[[email protected]]# id

uid=0(r) gid=0(root) groups=0(root)


I do manual pentesting and run every AI-assisted hacking tool I can get my hands on. Particular interest in agents that work with local models - if they don't support Ollama, I add it.

Ethics are fundamental, especially when you're breaking things.


Contact me:

GitHub

Email


Here's what website operators can see about you


Current Projects:

High-Focus

AI Pentesting Tools: A Running Survey [security] — live document, updated as I work through new tools

Labs

Hacking labs writeups [security]

2026

Vibe Coding and Git Secrets: What AI Assistants Won't Tell You [security]

Connecting Claude Code to Kali via MCP [security]

Installing Kali NetHunter on a OnePlus Nord [security]

2025

Installing Linux on an Intel iMac [hardware]

Automated Hacking Agents [security]

Open Source Intelligence in the age of MCP [opinion]

Gods and Lies: Where Should We Integrate? [opinion]

2024

Setting Up Open-Sora [development]

Glossary of Training Parameters [reference]

Setting up and using NVIDIA's garak framework on a local model [security]

Debugging nested JSON for NVIDIA's garak tool [security]

cuddles Setup and Usage Instructions [security]

AutogenGUI Install and Usage Instructions [development]

Scalable Autogen Frontend Binaries and Bonus Reversing [development]

Llama-3-70B-GGUF on an NVIDIA 3090 in 2-Bit Quantisation [performance]

Redteaming My Own Agent [security]

Quantisation On Local Hardware [performance]


> End of output.


foothold.php