How AI Agents Work
2026-08-01 · 6 min read
Models, tools, memory, and control flow — the four parts of every agent, explained without hype.
Four moving parts
- Model — the reasoning engine
- Tools — functions the model can call
- Memory — state carried between steps
- Control flow — what runs next, and when it stops
Most agent failures are control-flow failures
Agents rarely fail because the model is weak. They fail because nothing defined the exit, nothing validated the intermediate output, and nothing routed the failure anywhere useful.