• Home
  • Blog
  • Projects
  • Bookshelf
  • About

Whittling Code






Over the past few months I’ve been increasingly thinking about two differing strategies for building software;

  1. Shifting left, doing a lot of upfront design and specs on how the resulting software should look and work (think plan mode in coding agents)
  2. Starting with a rough shape of the desired result, and whittling/carving it down to the desired state

I used to largely be for the first of the above two strategies, and I think there are still times where it makes sense to pull forward some of the design and implementation discussions - especially when working with a team on a larger project.

However, AI agents have made it pretty easy to go with the second path, where it's far easier to "one-shot" a rough approximation of the thing you wanted with an agent, and then you can begin to away at the first result to get closer and closer to the desired product.

I don’t have a piece of software or a book to sell you on how you go about whittling, and I think it varies from person to person and harness to harness, but you should optimize your own workflow and setup for whittling away at the software.

Published
Last modified
From Boston, MA
....
....
← Back to all posts
Whittle
Share this post on Bluesky
See discussion on Bluesky
[1]Jump to footnote

For me, the workflow usually looks like the following:

  1. Give the agent a loose prompt, try to outline the desired goal from a light "user story" perspective, also nudge the agent towards approaching from a TDD sort of mindset - but with more emphasis on E2E style tests than unit tests
  2. Review the output, both by hand/human eye, and also via a "committee of agents" (usually a mixture of models and importantly in new threads/sessions to avoid context rot)
  3. Repeat step 2 above until the product looks good enough

Footnotes

↩Back to reference Maybe ironically, I think a lot of software engineers have worked to build up systems that use agents for the first path - anyone remember Spec-driven development from 2025??


Tags:

AIAgentic EngineeringSoftware Development

Loading...

Related Posts

AI

My AI Workflow - November 2026

....

A brief look at some recent experiments and changes to my software engineering workflow

Please try OpenAI models

....

OpenAI Enters the Agentic Knowledge Management Space

....

The Heat Death of Software

....

A few predictions about where software engineering is going, including some counter arguments as well

Your Job Is To Understand Systems

....

Parallel Computers

....

The Harness Isn't Your Problem, The Router Is

....

OpenBlank

....

ASD-STE100 Simplified Technical English

....

Performative Engineering

....

The act of making noise and delivering output instead of outcomes

The Agentic Engineering Scale

....

A brief outline of how I think of the scale of adoption of AI in software development

Onboarding Your New AI Teammate

....

Are we reinventing the wheel when it comes to onboarding new AI agents to a codebase, when we already have primitives available for onboarding humans?

Cursor and Obsidian

....

A quick look at combining Cursor with Obsidian to ask questions of your notes!

The AI Development Conundrum

....

Is it a good or a bad thing to offload writing code to AI agents and Large Language Models?

Offload Complexity, Don't Offload Learning

....

A rough mental model for how you should be leveraging AI as a tool for your own growth

My Current AI Stack

....

I've been using a variety of AI tools as of late, I figured I'd document the ones I'm primarily using!

Agentic Engineering

My AI Workflow - November 2026

....

A brief look at some recent experiments and changes to my software engineering workflow

The Eventually Consistent Codebase

....

Outlining how I now approach software engineering/development thanks to AI agents

The Heat Death of Software

....

A few predictions about where software engineering is going, including some counter arguments as well

Software Development

My AI Workflow - November 2026

....

A brief look at some recent experiments and changes to my software engineering workflow

The Eventually Consistent Codebase

....

Outlining how I now approach software engineering/development thanks to AI agents

The Heat Death of Software

....

A few predictions about where software engineering is going, including some counter arguments as well

Your Job Is To Understand Systems

....

Parallel Computers

....

The Agentic Engineering Scale

....

A brief outline of how I think of the scale of adoption of AI in software development

Your Job Is To Understand Systems

....

Parallel Computers

....

The Harness Isn't Your Problem, The Router Is

....

Performative Engineering

....

The act of making noise and delivering output instead of outcomes

The Agentic Engineering Scale

....

A brief outline of how I think of the scale of adoption of AI in software development

Onboarding Your New AI Teammate

....

Are we reinventing the wheel when it comes to onboarding new AI agents to a codebase, when we already have primitives available for onboarding humans?

Polish is Important

....

You're Building Software Wrong

....

Slicing software: why vertical is better than horizontal.