Frontend · Backend · WebXR · AI technologies
About me ↗
Six engineering principles - KISS, YAGNI, DRY, AHA, SoC, and SOLID - without the cult. On one running example (an online-store order calculation) you see where a rule saves the code and where it turns into new complexity. A principle is useful while it helps you think.
When several agents actually pay off and when they just burn budget. Router, supervisor, agents-as-tools, and handoff; context and write isolation, limits, failure resilience, trust boundaries, and proving usefulness on your own tasks - across the OpenAI Agents SDK, LangGraph, AutoGen, and CrewAI.
The engineering discipline of working with coding agents: a task contract, repository research, a checkable plan, a minimal patch, tests in a bounded environment, and independent verification. Codex, Claude Code, and Cursor are covered on the same task.
How to assemble a specialized AI agent on providers' direct APIs rather than a framework. A running online-store support agent lives as a provider-agnostic business contract and works the same on OpenAI Responses, Anthropic Messages, and Google Gemini Interactions.
Two connectivity protocols for AI systems under one cover: MCP wires up capabilities and data, A2A links independent agents. It all runs on a Release Review Agent - a real TypeScript project on SDK v2, proven by 14 tests with no API keys.
How to design instructions, context, and control for RAG, support, call centers, assistants, orchestrators, and agents - without jailbreaks or spellbooks that outlived the models they were written for.