What is harness engineering?

Yeah.

So that was, uh, one of the challenges when I was writing the article to figure out how to even define it, because also in these, uh, days of AI, people, uh, create a lot of content and there's, like, a lot of discourse happening, like, very, very quickly.

There's a lot of throughput of communication as well, right? So we just throw out a lot of terms and, uh, and, and then get semantic diffusion really quickly.

So I, I ended up kind of describing it almost like an onion kind of model.

So you have the, the large language model, uh, as your ultimate tool that you use to do something, but then you, uh, put something around it, uh, that, that people have started calling the har- the harness, right? And so a coding agent is a harness of an LLM in a way.

So Claude Code is a harness, um, uh, the Pi coding agent is a harness, right? So there's lots of choices there.

And the, the way that they're harnessing is by putting together a bunch of tools that can be, uh, that can be used by the, the LLM through this harness to do stuff, right? So in the case of coding agents, that is like editing files, writing fi- uh, uh, reading files, uh, certain code search tools, right? Like maybe access to a language server, all of that type of stuff, right? And then they also orchestrate prompts.

They have a system prompt.

So all of that is kind of like making the model much more useful to, to be able to code for us, right? But then as coding agent users, as users of this harness, we can also expand the harness, so that's kind of like the next onion layer out, right? So we can take, uh, the ready-made harness like Claude Code or Cursor that, you know, some people have already thought very deeply about, like, what are all the things we need for coding, but then make it more specific to what we are working on, right? So if I work on a TypeScript, uh, code base, um, I wanna think about, like, how-- like, what are specific things in my particular application and TypeScript code base that I, um, also want to harness, right? So for example, you know, we'll get into it later, like static code analysis or, or stuff like that, or what are additional tools I wanna make available to it? What are my guidelines, my, uh, specifications and stuff like that that I feed into it? And so that's also what people have started calling a harness, right? I almost wish it was a different word because [chuckles] that might make the, make it easier.

It's almost like two different bounded contexts, right? And so that's what people now call harness engineering, and I would always, like, when, when you have conversations with other people about this, I would always try to make sure that everybody's talking about the same type of harness so that there's no misunderstandings.

[chuckles]