Joël Quenneville
Principal Developer, thoughtbot
Joël has been writing Ruby for over a decade and is a fan of the language’s focus on developer happiness and productivity. He co-hosts the podcast “The Bikeshed” where he discusses the technical and social issues that surround our industry.
Harness Engineering on Rails
Agents have re-shaped how AI can work with code yet we still use them like ChatGPT and wonder why they disappoint.
The quest to one-shot our code with the perfect prompt or AGENTS.md is a trap. We get much better results by hooking into signals from RSpec, Rubocop, or even a self-review. Our agent can react to these and iterate to a good solution before requiring a human in the loop. This requires a paradigm shift on what an agent is: from chatbot to reactive system. Building on this perspective we’ll explore techniques to keep our agents on the rails, prevent their ability from decaying, and extend their capabilities.
When your reaction to mediocre code is to fix the harness rather than the mistake, agents can become valuable assistants that write code you are proud of.