Raw LLM Responses

Inspect the exact model output for any coded comment.

Comment
I currently write platform code for enterprise software. Claude is doing all of the boiler plate stuff and I spend my time reviewing it and doing the finishing touches like error handling, environment config setup, monitoring and alerting, etc. Then it gets reviewed again by another developer. I save a lot of time not writing all the code by hand. I've been a software developer for 16 years so I know how to look for the patterns that will cause bugs or don't work at scale pretty quickly.  The other crazy thing is that it's only going to get better. And now Claude is so integrated into everything like Jira, Datadog, Chrome, etc that I can fully automate certain repititive things. I've done a full 180 after the last couple months seeing what it's capable of.
reddit AI Jobs 1772499468.0 ♥ 4
Coding Result
DimensionValue
Responsibilitynone
Reasoningconsequentialist
Policynone
Emotionapproval
Coded at2026-04-25T08:33:43.502452
Raw LLM Response
[ {"id":"rdc_o89sr9c","responsibility":"company","reasoning":"consequentialist","policy":"none","emotion":"resignation"}, {"id":"rdc_o8bezrk","responsibility":"company","reasoning":"consequentialist","policy":"none","emotion":"indifference"}, {"id":"rdc_o8c7nyk","responsibility":"company","reasoning":"consequentialist","policy":"none","emotion":"outrage"}, {"id":"rdc_o8bwzou","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"}, {"id":"rdc_o8b3101","responsibility":"none","reasoning":"consequentialist","policy":"none","emotion":"approval"} ]