SERIES 02Twelve essays · Harness Engineering

Build theagent,not the model.

An illustrated field guide to the harness — the system wrapped around a model that turns it into an agent that ships. Environments, memory, tools, planning, security, and the loop that improves itself.

12
Illustrated essays
296
Minutes of reading
65k
Words, fully sourced
4
Parts, one argument

The Essays

Four parts · read in order
AGAgents & HarnessARAgentic RLCLContinual LearningMEMemoryTOTools & SkillsMAMulti-AgentSWSoftware EngineeringEVEvaluationSESecurityFNFoundations
Series 03 · The Long BetWhat survives in agentic AIThe long-horizon companion to this field guide — a selection lens for the literature and the three-to-ten-year bets it picks. Six illustrated essays.Read the series →

“Two teams deploy the same frontier model. One ships an agent that resolves most of its tasks; the other stalls, loops, and quietly fails. Nothing about the model varied.”

This series follows a single thread: a deployed agent's capability is the product of model quality and harness quality — and in 2026 the harness has the steeper gradient. Each essay pairs a close read of the primary literature with a picture you can hold in your head, and ends in something you can build.

Harness Engineering · the sister series to Continual Intelligence