Design, build, and deploy autonomous AI agents that get real work done.
Cut through the marketing. Define agents by behavior, not hype.
How tool calling works under the hood, and how to design tools models can use.
The three kinds of memory an agent needs and how to build each.
Different shapes of agent reasoning and when to use each.
When multiple agents help, when they don't, and how to coordinate them.
Why agent eval is different from LLM eval, and the harness patterns that work.
Defense in depth for agents that take real actions.
Ship a working research agent with tools, memory, and eval.