[BUG] Init Feature Doesn't Do Much, If Anything

Preflight Checklist

• [x] I have searched existing issues and this hasn't been reported yet
• [x] This is a single bug report (please file separate reports for different bugs)
• [x] I am using the latest version of Claude Code

What's Wrong?

The /init feature, while a wonderful idea, is hardly useful and might actually do nothing at all. The "Comprehensive" report it creates it pretty much useless - outlining folder structure and summarizing the project. I'm pretty sure the agent can see the folder structure the second it begins, it's what lies inside of the folders and even better: the files, that really matters.

What Should Happen?

I would like to see the agent checking the /init report each time it begins and again when it ends ends a conversation. Once to check and another to make notes about anything it might have learned.

Learned:

• Whats in the files

• Important line ranges and what function is there

• Previous errors, break fixes etc

• Maybe even notes about the skill level of the user could be helpful. The agent could know to ask questions instead of assuming.

• Lots of potential with an empty slate going to waste here.

Init reports should be free too, not counted against rate limits.

Error Messages/Logs

Steps to Reproduce

run /init

Claude Model

None

Is this a regression?

I don't know

Last Working Version

_No response_

Claude Code Version

Latest

Platform

Claude Code CLI

Operating System

Windows

Terminal/Shell

PowerShell

Additional Information

_No response_