[BUG] MultiEdit UTF-8 Encoding Bug Report

[dascwf]

Environment

• Platform (select one):
  • [ ] Anthropic API
  • [ ] AWS Bedrock
  • [ ] Google Vertex AI
  • [x] Other: Claude Code CLI with MultiEdit tool
• Claude CLI version: 1.0.24 (Claude Code)
• Operating System: Linux (root@build) Alpine Linux ^3.22.0
• Terminal: Fish Shell 4.0.2+

Bug Description

The MultiEdit tool incorrectly handles UTF-8 encoded content when writing files, resulting in corrupted characters (mojibake). Chinese characters and special UTF-8 symbols are replaced with replacement characters (�) in the written files, while the same content displays correctly in the Claude interface.

Steps to Reproduce

1. Use Claude Code to create a new markdown file with Chinese content:

   MultiEdit prompt/skeleton 3.1.3.md
   

2. Include Chinese characters in the content, such as:
   • "骨架屏" (skeleton screen)
   • "中文文档" (Chinese documentation)
   • Special characters like "—" (em dash) or "•" (bullet)
3. Save the file using MultiEdit
4. Read the file back using:

   Read prompt/skeleton 3.1.3.md
   

Expected Behavior

• File should be saved with proper UTF-8 encoding
• Chinese characters and special symbols should be preserved
• Reading the file should display the same content that was written

Actual Behavior

• Chinese characters are replaced with � (replacement character)
• Special UTF-8 characters are corrupted
• File content shows mojibake when read back
• Example: "骨架屏" becomes "��架��"

Additional Context

Test Case Evidence:

• Claude interface displays: "骨架屏（Skeleton Screen）"
• File contains: "��架��（Skeleton Screen）"

System Encoding:

$ echo $LANG
en_US.UTF-8

File Encoding Check:

$ file -bi "prompt/skeleton 3.1.3.md"
text/plain; charset=utf-8

Workaround: Currently avoiding Chinese characters and using only ASCII in documentation files.

Impact: This bug prevents creating multilingual documentation or using special UTF-8 characters through Claude Code's MultiEdit tool, limiting its usefulness for international users and technical documentation that requires special symbols.

Suggested Fix: Ensure MultiEdit tool explicitly uses UTF-8 encoding when writing files, possibly by:

• Setting encoding explicitly in file write operations
• Using Buffer.from(content, 'utf8') when handling string content
• Ensuring all string transformations preserve UTF-8 encoding

[eefan000]

I encountered the same problem, when using Rider, the claude code was still correct when using IDE's difference comparison function, but when I clicked accept, the encoding was incorrect! This issue greatly affects usage and is equivalent to wasting a conversation! Please fix it as soon as possible!!!

[aifirstd3v]

MacOS, Vscode and

❯locale 
LANG="en_US.UTF-8"
LC_COLLATE="en_US.UTF-8"
LC_CTYPE="en_US.UTF-8"
LC_MESSAGES="en_US.UTF-8"
LC_MONETARY="en_US.UTF-8"
LC_NUMERIC="en_US.UTF-8"
LC_TIME="en_US.UTF-8"
LC_ALL="en_US.UTF-8"

Whenever claudecode writes markdown content, It always writes like this: "�� Workers-Hono (Cloudflare), � Pending:, � Pending:"

I set up corretly locales, Lang stuff. Other AI CLI tools have no issue but only Claudecode.

[tjx666]

MacOS cursor:

[CorrectRoadH]

I got same issue

[github-actions[bot]]

Found 3 possible duplicate issues:

1. https://github.com/anthropics/claude-code/issues/1716
2. https://github.com/anthropics/claude-code/issues/5047
3. https://github.com/anthropics/claude-code/issues/5111

This issue will be automatically closed as a duplicate in 3 days.

• If your issue is a duplicate, please close it and 👍 the existing issue instead
• To prevent auto-closure, add a comment or 👎 this comment

🤖 Generated with Claude Code

[github-actions[bot]]

This issue has been automatically closed as a duplicate of #1716.

If this is incorrect, please re-open this issue or create a new one.

🤖 Generated with Claude Code

[github-actions[bot]]

This issue has been automatically locked since it was closed and has not had any activity for 7 days. If you're experiencing a similar issue, please file a new issue and reference this one if it's relevant.