AX
Agent experience: how well the environment is set up for an agent to do good work.
AX is agent experience: how well the environment is set up for an agent to do good work. Where DX asks whether a human can work well here, AX asks the same about an agent, and the answers differ in instructive ways.
An agent needs a few things a human takes for granted: a way to find out what is true without guessing, a way to verify its own work, clear boundaries it will not cross, and instructions that do not contradict each other. It does not need a nice onboarding flow, but it does need a machine-readable one. It does not read a comment, but it does read a well-named file.
The gap between DX and AX is where most of the easy wins are. A repository with a slow test suite, a confusing directory layout, and error messages that do not say where the problem is has good DX for nobody. Fixing it for the agent usually improves it for the human as well, because both want the same things: to know what is true quickly and to find out whether the change worked.
Related terms
DX
Developer experience: how easy a codebase and its toolchain make it for humans to do good work.
AGENTS.md
A file in the environment that the harness loads into the context at session start: the project's standing brief to the agent.
Testing
The cheapest way to know a change is safe: a check that runs without a human reading anything.
Environment
The world the agent acts on: anything outside the harness that it perceives via tool results and changes via tool calls.
Context Engineering
Deciding what goes into the model's context, and what stays out.
Want this working inside your team?
It starts with one conversation about your workflow.