Writing
Notes on web infrastructure, hosting, and the tools that keep sites online. Write-ups, breakdowns, and the occasional post-mortem.
Follow on X.
2026
Why agent observability Claude failures security demands full trace capture: tool calls, reasoning, and retrieved context for real incident response.
How AI is reshaping AI and WordPress security in 2025, from faster exploits to smarter host-side detection. Here's what site owners need to know.
Perplexity hybrid compute Mac explained: how the on-device PII-Tracer privacy gate works with local models for secure AI inference.
Learn how to read the Google Search Console AI performance report and opt your site out of Google AI Overviews with the new toggle.
Anthropic Model Hardware Standard MHS preview cuts lab instrument integration from weeks to hours with a shared driver spec for AI agents.
Google Search Console AI performance report is now global. See how the new AI filter and generative AI traffic segment change SEO reporting.
Three SEO KPIs tied to one goal beat 18 on a dashboard. How to choose them, where the data comes from, and why AI visibility scores need a caveat.
AI agents in developer platform work: real roles, MCP integration, identity risks, and what to expect next for platform teams.
Cosmos EVM vulnerability exploit 2026 drained six chains after a silent patch. What GHSA-7g4w-cg88-2cq2 means for operators and forks.
LLM agents for incident response are reshaping SOCs. Can they block IPs automatically? A practitioner's take on blast radius, reversibility, and guardrails