~/blogs/security
$ls --sort=latest|
Security research essays on vulnerabilities, exploitability and AI agent safety.
security2026-08-1716 min read
I Asked an AI Agent to Go Live. It Exposed My Server.
A live test of DeepSeek Harness, a self-hosted AI coding agent, shows what happens when an obedient agent is asked to expose a server, and why the firewall around it mattered more than the model inside it.
#security#ai-agents#infrastructure#deepseek
cat post.mdsecurity2026-07-1813 min read
CVE-2026-63030: WordPress Pre-Auth RCE Explained
CVE-2026-63030 is critical, but automated exploitation fails on one overlooked detail that 70% of production sites already defend against.
#wordpress#cve-2026-63030#security#exploitation
cat post.md