Agent engineering.
Subagents, skills, MCP servers, RAG — the building blocks of agents that do real work, each explained with something you can build.
Swap the LLM, keep the code: one RAG app on Azure, OpenAI and local Mistral
The application code never changed when I swapped the LLM three times. The configuration nearly beat me — and the two traps are worth knowing.
Keep readingRAG explained by building it: the smallest working app with Spring AI
You don't need LangChain or a vector database to understand RAG. The whole loop fits in two Java classes — and building it that small is the point.
Keep readingAutomate invoicing with MCP: my Invoice Ninja server
“Log 45 minutes of weekly meeting on the Audioempire project, cost 0.” One sentence, done. Behind it sits an MCP server I run myself — and one day where it deleted tracked hours.
Keep readingClaude Code Subagents vs. Skills: The Practical Guide
An agent that does everything does nothing well. Subagents and skills are the difference between a toy and a tool — here you build both, from a real example, and learn when each abstraction is the right one.
Keep readingMCP Enterprise Adoption: Outgrowing the Tinkering Phase
97 million SDK downloads, enterprise auth in the standard — MCP has outgrown the laptop. What that means for teams moving from prototype to production, and the five questions to answer before the first production server.
Keep reading1M Token Context: Feed Your Agent Like an Architect
Sonnet 5 reads a million tokens. Most people still feed their agent one file at a time — and give away the actual leap. What big context changes in practice, where it hurts, and a rule of thumb for daily work.
Keep readingWrite your first MCP server in 30 minutes
An MCP server isn't magic: a handful of functions, cleanly described. Here you build your first one and plug it into your coding agent.
Keep readingAI code without tech debt — the checklist
Sign up: the checklist plus new posts on AI engineering. No spam, unsubscribe anytime.