[BUG] Rust panic when parsing CLAUDE.md with Korean (multi-byte UTF-8) characters

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?

Bug Description

Claude Code crashes with a Rust panic during init when CLAUDE.md contains Korean (or other multi-byte UTF-8) characters.

## Error Message

thread '' panicked at /rustc/.../library/core/src/str/mod.rs:833:21:
byte index 10 is not a char boundary; it is inside '와' (bytes 8..11) of ChatGPT와 Gemini에 보내는 크롬 익스텐션(Manifest v3)

## Steps to Reproduce

1. Create a CLAUDE.md file with Korean text (or other multi-byte UTF-8 characters)
2. Run claude to init in that directory
3. Claude Code crashes immediately

## Expected Behavior

CLAUDE.md with non-ASCII UTF-8 characters should be parsed correctly without crashing.

## Root Cause

The Rust code is slicing a string at a byte offset (10) that falls in the middle of a multi-byte UTF-8 character ('와' spans bytes 8-11). String slicing in Rust must occur at valid character boundaries.

## Workaround

Temporarily rename or remove CLAUDE.md before running init, then restore it afterward.

## Environment

• OS: Windows
• Claude Code version: (latest)

What Should Happen?

CLAUDE.md files containing Korean (or other multi-byte UTF-8) characters should be parsed correctly during claude init without crashing.

Error Messages/Logs

Steps to Reproduce

1. Create a CLAUDE.md file with Korean (multi-byte UTF-8) content, for example:

```markdown
# CLAUDE.md

Project Name - ChatGPT와 Gemini에 보내는 크롬 익스텐션

1. Run claude to initialize in that directory
2. Claude Code crashes with Rust panic:

thread '<unnamed>' panicked at .../library/core/src/str/mod.rs:833:21:
byte index 10 is not a char boundary; it is inside '와' (bytes 8..11)

Note: The bug occurs when Rust code slices a UTF-8 string at a byte offset that falls in the middle of a multi-byte character (Korean characters are 3 bytes each in UTF-8).

Claude Model

None

Is this a regression?

Yes, this worked in a previous version

Last Working Version

_No response_

Claude Code Version

 2.0.76 (Claude Code)

Platform

Anthropic API

Operating System

macOS

Terminal/Shell

Terminal.app (macOS)

Additional Information

_No response_