The Harness Is the Thing: How One Developer Cut Frontier Model Costs by 75%
Scott Fryxell describes how he built a personal AI development harness that unifies multiple TUIs (Cursor, Claude, and Pi) around shared skills and AGENTS.md. By using a planner/worker/critic/promoter workflow and reserving frontier models for planning and promotion, he cut frontier usage by 75%, relying on cheaper models like DeepSeek for most tasks. The harness also extends to his product, enabling advanced use cases like scripting the app via a headless browser. He argues that the harness—not the model—is the key to productivity, and that commoditized models make switching trivial.
The harness is the thing; the fulcrum from which my expectations meet the LLM's capabilities.
- AirMax98
Reading this really makes me wish that I had a slightly better workflow. I'm really soley dependent of Fable to the point that I don't use other models, and I've already sort of hit a point where I'm running into usage limits every week. I am really living on borrowed time — when Anthropic finally collapses their 50% usage increase at the end of August, I'll definitely be forced to switch my workflow. When that happens, I have a hard time imagining that I'll be sticking with a single model on a single provider.
- andai
I've been very happy with Luna but my approach is "many bite sized edits" for which models basically hit saturation a year ago.
(I also tried the "let a massive model make massive changes" approach and am still psychologically recovering from the experience. The codebase may never recover!)
Also, Luna and DSV4 Flash seem to be on par now except Luna is faster and cheaper?
- esalman
What I've learned in last week is that a harness is basically a while loop.
In each iteration you make an LLM call, perform some work (e.g. tool call), augment the prompt (append or compact etc.)- not necessarily in that other- and continue.
Until an end condition is satisfied. Then you break out.
- nullbio
LoRA adaptive learning using open-weight models and your own reasoning traces is the thing. The big labs have a mammoth job ahead of them if they want to compete with running your own model - they will basically have to give every single user their own persistent virtual machine. When it's all said and done, I think their only really moat will be as inference/hardware providers. Stripe buying OpenRouter was a very smart bet.
- brainless
I use Claude Code, Codex and opencode pretty much interchangeably. I am currently using Claude more this month because (stupidly) I paid for Max ($100) since I have a large client project.
I generally use larger models to plan. All my generated Epics have similar structure. All my repos have similar structure (https://github.com/brainless/akar and https://github.com/brainless/daftprompt are recent examples).
I barely spend time or thought in making prompts. I have a simple text file with a few combinations. They refer all the common files (README, AGENTS, DEVELOP, etc.)
All reference software is cloned locally and the docs mention that. The prompt templates then boil down to research mode (write Epic) or worker mode (write software) or review mode (leave review notes in Epic). That's it.
Many of my harness experiments are about text manipulation, text search, graph on text. Because that is what LLMs are - text processing systems. Cut parts of prompts, cut parts of response, cut parts of user's intent. Join, break into epics/tasks, run with LLMs, repeat.