@ file search is almost unusable in huge code base

Summary

Use "@" to start a file search, it will almost always freeze in large code base. This issue lasts for a long time since I start to use cmdc.

I tested this in two repos, one of it has 42906 files and one of it has 109765 files, the "@" search works but delay like 3-5 seconds on the 42906 one, and will get stuck on the 109765 one.

It seems the solution to file search is not properly implemented in this tool. Switch to codex or claude-code, both of them works seamlessly with "@" file search. Also I believe the current implementation does not ignore some common dependency paths or cache folders.

Expected Behavior

File search works and should be fast

Actual Behavior

File search gets really slow with large code base, and will stuck with huge code base.

Steps to reproduce the issue

Try check out a big mono repo, and install the dependencies fully, then run the file search worfflow.

Command Code Version

1.32.1

Operating System

Linux

Terminal/IDE

No response

Shell

zsh

Session file (optional)

No response

Fix prompt (optional)

No response

Additional context

No response

No file or test is identified in the issue. Reproduce the @ file-search workflow on Linux with zsh in repositories containing 42,906 and 109,765 files, with dependencies installed, then trace the existing search entry point. Done means the search stays responsive without the reported multi-second delay or freeze.