Build an MCP server people actually use
These are the parts I usually build, and the traps that decide whether the server gets used or stays in a folder.
Practical guides to MCP, agents, documents, local models and how to test AI before it meets customers. No shortcuts and no promises of magic.
All guides
These are the parts I usually build, and the traps that decide whether the server gets used or stays in a folder.
The hard part is not getting the model to call a tool. It is getting it to stop in time, do the right thing when something fails and leave a trail behind.
RAG is less about the model than about which documents are found, who may see them and how the answer shows where it came from.
Own hardware solves data protection and sometimes cost. It does not solve quality, updates or who is responsible at three in the morning.
AI is most useful at the edge of a flow: reading, interpreting and suggesting. The rest should be ordinary, predictable code with clear checks.
A good AI system has a set of cases you can run again and again. Without it, you only know it worked on the day someone happened to look.
AI assistant
Answers from what is on this site
The chat uses AI. Please don’t share personal details here.