Harness Engineering: Boosting AI Agent Output by 100x Through Context
I believe most people don't realize they can point their agents at my writing and improve output by 100x. Harness engineering treats the model as a black box, focusing instead on shaping the environment with context and tools. By embedding an organization's nonfunctional requirements into code, we make judgment cumulative. This approach ensures agents recover intent, respect authority, and leave systems better equipped for future runs.
Most people do not know that they can just point their agents at my writing, tweets, podcasts, and talks and improve the output of their agents by 100x.
- ricardobeat
What percentage of this was written with AI?
> A command can feel ambient without making its credential model-visible.
I'd be wary of following guidelines on how to build human-computer interaction systems that weren't written with full human oversight. This kind of wording makes me wonder if these are actual recommendations or just the AI pattern-matching and hallucinating something that looks coherent.
- lopopolo
Hi folks, author here and also author of the seminal OpenAI blog on this topic. Let me know how I can help you all let it rip.
- turing_complete
Starting a repo readme with a quote of yourself...
- K3UL
Prompt engineering, context engineering, now harness engineering. I guess in a few months we'll have looped back to software engineering ?
All these buzzwords for basically what's glorified tweaking and configuration. All these years I should have mentioned I was doing Debugger Engineering and IDE Engineering
- richardfey
Serious question: how do we verify claims like these on the effectiveness of a harness?