AI Agents · Security
Building Secure AI Agents with Tool Calling
How to design AI agents that use tools safely — grounding, bounded loops, trust levels, prompt injection defense, and audit trails for production security agents.
June 2026 · 14 min read
Post-Quantum Cryptography
Replacing X25519 with ML-KEM in WireGuard
WireGuard's elegant handshake relies on X25519 — vulnerable to quantum attack. A deep dive into integrating NIST's ML-KEM via a hybrid KEM approach, with benchmarks on ARM hardware.
June 2026 · 12 min read
Coming Soon
How I Built an AI Agent for Network Security
End-to-end walkthrough of building SecureAI Agent — ingesting logs, PCAP, and CVEs with OpenAI tool calling to produce analyst-grade threat reports.
Coming July 2026 · ~12 min read
Coming Soon
Deploying AI in Air-Gapped Networks
Lessons from building AirGapGPT — running LLMs, RAG, and RBAC in environments with zero internet access.
Coming July 2026 · ~10 min read
Coming Soon
Can LLMs Understand PCAP Files?
An empirical look at feeding raw packet captures to language models — what they understand, where they fail, and how to structure PCAP data for LLM consumption.
Coming August 2026 · ~10 min read