A staff engineer's journey with Claude Code

Preventing garbage just requires that you take into account the cognitive limits of the agent. For example ...

1) Don't ask for large / complex change. Ask for a plan but ask it to implement the plan in small steps and ask the model to test each step before starting the next.

2) For really complex steps, ask the model to write code to visualize the problem and solution.

3) If the model fails on a given step, ask it to add logging to the code, save the logs, run the tests and the review the logs to determine what went wrong. Do this repeatedly until the step works well.

4) Ask the model to look at your existing code and determine how it was designed to implement a task. Some times the model will put all of the changes in one file but your code has a cleaner design the model doesn't take into account.

I've seen other people blog about their tricks and tips. I do still see garbage results but not as high as 95%.

At that point, why not just write the code yourself?

I reached this conclusion pretty quickly. With all the hand holding I can write it faster - and it’s not bragging, almost anyone experienced here could do the same.

Writing the code is the fast and easy part once you know what you want to do. I use AI as a rubber duck to shorten that cycle, then write it myself.

I’ve been trapped in a hole of “can I get the agent to do this?” And the change would have taken me 1/10th the time.

Choosing the battles to pick is part of the skill at the moment.

I use AI for a lot of boiler plate, tedious tasks I can’t quite do a vim recording for, small targeted scripts.

A similar, non-LLM battle, is a global find and replace, but _not quite identical_ everywhere. Do I just go through the 20 files and do it myself, or try to get clever with regex? Which is ultimately faster...