Writing
Make the invisible understandable. Technical writing on AI agents, developer tooling and observability.
Articles
-
Back to Basics: What Joel Spolsky Said in 2001 Still Decides Who Can Use AI Well
AI is the thickest abstraction we have built, and abstractions leak. Why memory, complexity, concurrency and reading code matter more now, not less.
Nov 2026 · Medium · Software engineering, AI, Computer science -
AI: From Zero to Hero
Seventy years of AI in ten milestones, the move from chat to agents, what tools, MCP, connectors, skills and plugins are, and a first automation you can build in fifteen minutes.
Nov 2026 · Medium · AI agents, Claude Code -
No-AI Days: Use AI Heavily, but Keep Your Brain in the Gym
Why I use AI hard every day and still keep one No-AI day a week. Brains are like muscles, and the skill you do not practice fades.
Oct 2026 · Medium · AI, Software engineering -
Harness Engineering in Under 3 Mins: How to Make AI Coding Agents Reliable
What harness engineering is, the agent loop, and six rules that make AI coding agents reliable. The full version is a Deep Dive on Medium.
Oct 2026 · serkan.fyi · AI agents, Developer tools -
What Your AI Assistant Actually Receives, and How to Keep Personal Data Out of It
What leaves your machine when you paste a record into Claude or ChatGPT, and how mask2ai replaces personal data with placeholders before it goes.
Sep 2026 · Medium · Privacy, AI, mask2ai -
Tokenmeter: Know What Every Prompt Costs You in Claude Code, Codex and Copilot
Why token counts hide the real cost of a prompt, and how tokenmeter reads your agents' own logs to show it.
Sep 2026 · Medium · Developer tools, AI, tokenmeter -
Capturing Your Screenshots with Foresight in Action
A practical walkthrough of test screenshots and CI visibility with Thundra Foresight.
Dec 2021 · Medium · Observability, Testing, Java
Originally published on the Thundra blog in December 2021. The Medium edition preserves the original walkthrough; product interfaces may have changed since publication.