←back to thread

A staff engineer's journey with Claude Code

(www.sanity.io)
548 points kmelve | 1 comments | 02 Sep 25 19:34 UTC | HN request time: 0s | source
Show context
swframe2 ◴[02 Sep 25 20:56 UTC] No.45108930[source]▶
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%.

replies(20): >>45109085 #>>45109229 #>>45109255 #>>45109297 #>>45109350 #>>45109631 #>>45109684 #>>45109710 #>>45109743 #>>45109822 #>>45109969 #>>45110014 #>>45110639 #>>45110707 #>>45110868 #>>45111654 #>>45112029 #>>45112178 #>>45112219 #>>45112752 #
dontlaugh ◴[02 Sep 25 22:30 UTC] No.45109969[source]▶
At that point, why not just write the code yourself?
replies(4): >>45110017 #>>45110032 #>>45110323 #>>45111832 #
lucasyvas ◴[02 Sep 25 22:36 UTC] No.45110017[source]▶
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.

replies(3): >>45110095 #>>45110162 #>>45112850 #
2muchcoffeeman ◴[02 Sep 25 22:46 UTC] No.45110095[source]▶
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.

replies(2): >>45110722 #>>45111616 #
skydhash ◴[03 Sep 25 00:05 UTC] No.45110722[source]▶
How many of these boilerplate do you actually have to do? Any script or complicated command that I had to write was worthy to be recorded in some bash alias or preserved somewhere. But they mostly live in my bash history or right next to the project.

The boilerplate argument is becoming quite old.

replies(2): >>45111525 #>>45111668 #
1. 2muchcoffeeman ◴[03 Sep 25 02:27 UTC] No.45111668{3}[source]▶
I’m doing a lot of upgrades to neglected projects at the moment and I often need to do the same config over and over to multiple projects. I guess I could write a script, or get AI to write a script, but there’s no standard between projects. So I need the same thing over and over but from slightly different starting points.

I think you need to imagine all the things you could be doing with LLMs.

For me the biggest thing is so many tedious things are now unlocked. Refactors that are just slightly beyond the IDE, checking your config (the number of typos it’s picked up that could take me hours because eyes can be stupid), data processing that’s similar to what you have done before but different enough to be annoying.