Complex systems research — security architecture, artificial intelligence, social dynamics and applied cognitive psychology. Each article traces a problem from its technical layers to its human and systemic implications.
The MCP OAuth phishing attack exploits VS Code's automatic authentication flow. No vulnerabilities are required: the protocol works exactly as designed. Analysis of the mcp-phish-poc, remote exploitation, and the cascade of MCP security findings in June 2026: Axios hijack, mcp-remote RCE, Claude Code token theft, CVE-2026-41613, and the GitHub breach by TeamPCP.
A 40-page forensic analysis of the H.M. case (Calama, 2026) — the school attack that triggered a regional contagion effect. Cinthia Gonzalez Castro's investigation dissects how silent emotional suffering, amplified by recommendation algorithms and digital communities of nihilistic violence, transforms into an armed identity. From the linguistic architecture of the blackpill to an interactive evidence board: a framework for understanding — and dismantling — the conditions that produce adolescent digital radicalization.
Google GTIG published a report claiming to have detected the first AI-generated zero-day. The evidence is thinner than it seems, and this isn't the first time this claim has been made. An analysis of the markers, the precedents, and what was left unsaid.
How Model Context Protocol can orchestrate autonomous red team agents across complex network environments. An analysis of protocol design, agent coordination patterns, operational memory and the security implications of self-directed offensive systems.
More articles in progress
Architecture Notes often evolve into full articles. Follow the research pipeline from Architecture Notes →
Get expert insights on cybersecurity, AI systems architecture, and strategic technology consulting — delivered to your inbox. No spam, ever.
No spam · Unsubscribe anytime