The AI words, explained properly
People use these words as if everyone already agreed what they mean. Here is what each one actually is, in basic English, with a drawing and a real example. Nothing here assumes you can code. 11 explanations, free to read.
How agents are built
The words people use for how AI systems are put together.
- Prompt Engineering9 min
Writing the instruction so the answer comes back right
- Context Engineering10 min
Choosing what the AI gets to see while it answers
- Loop Engineering10 min
Teaching an AI to check on its own, no prompt needed
- Graph Engineering10 min
How many AI agents connect and work as a team.
- The Agent Harness11 min
The system built around the AI that makes it actually useful
- Pi: The Bare-Bones Toolkit Behind Popular AI Coding Agents7 min
A stripped-down AI toolkit that other AI tools are built on top of.
Running them for real
What happens after it works once: facts, records, tests and safe fixes.
- RAG: Giving the AI Your Own Facts10 min
Look it up first, then answer
- Tracing: Seeing What the AI Actually Did7 min
One record per run, so you can find the broken step
- Evals: Checking If It Is Any Good8 min
A test you can run again, instead of a feeling
- LLM Ops: Keeping It Working9 min
Building it is day one. This is every day after.
Set one up yourself
Start to finish walkthroughs, for when reading about it is not enough.
- Hermes Setup12 min
Your own AI assistant, running 24/7 and messaging you on Telegram