# Not reading AI code line by line, but not skipping it either. I sort it by "does it hurt if this is wrong" and "can I follow it". How do you decide what to actually read? — retained record

> Retained reddit record from reddit:AI_Agents.

Canonical: https://highsignal.app/data/records/0b79df79c284fd3c

[← sources](https://highsignal.app/data)


source: [reddit](https://highsignal.app/data/reddit) (reddit:AI_Agents)  published 2026-10-13 09:55 ingested 2026-10-14 02:36

[original source ↗](https://www.reddit.com/r/AI_Agents/comments/1x4shph/not_reading_ai_code_line_by_line_but_not_skipping/)

## retained text



```
Been going back and forth on the "should you read the code the AI writes" question, especially for people who are new and honestly can't follow most of it yet.

Reading every line isn't realistic if you don't know what `computed` or `ref` means. But skipping it entirely feels like driving with no dashboard. You have no idea how fast you're going until someone pulls you over and asks.

What's been working for me is asking two questions about each piece of code:

1. If this is wrong, does it hurt? (wrong amounts, lost data, deleting the wrong thing)
2. Can I actually follow it?

That gives four boxes:

- Hurts if wrong, can follow it: read it yourself once. Money math, discount logic.
- Hurts if wrong, can't follow it: make the AI explain it until you can. Where the data is stored, how it's saved and loaded.
- Doesn't hurt, can follow it: a quick skim is enough. Button labels, message wording, colors.
- Doesn't hurt, can't follow it: leave it to tests and code review. Style setup, library init, config files.

The bar I use for "read" is low on purpose. Can you explain in plain words where the data lives, what shape it has, and which files it passes through between tapping "add an expense" and the screen updating? Not every line, just the path.

Also, the line moves with the project. For a throwaway prototype I barely bother. For something other people will use, the "hurts if wrong" row gets a lot bigger.

How do you all decide what to actually read? Do you have a similar line, or do you lean on tests for everything?
```



document fetched 2026-10-14 00:19 · canonical https://reddit.com/r/AI_Agents/comments/1x4shph/not_reading_ai_code_line_by_line_but_not_skipping
