---

name: "cn-punctuation-checker" description: "Checks Chinese text for incorrect English punctuation marks. Invoke when user wants to find and fix wrong punctuation in Chinese copy."

Chinese Punctuation Checker

A skill for detecting incorrect English punctuation marks in Chinese text.

Features

1. Auto Detection: Detect English punctuation marks incorrectly used in Chinese text
2. Precise Location: Report line number, column number, and context snippet for each error
3. Report Generation: Output Markdown format inspection report
4. Batch Fix: One-click replacement of all incorrect punctuation with correct Chinese punctuation
5. Project-wide Scan: Support scanning entire project with automatic exclusion of code files

Supported Punctuation Marks

English	Chinese
,	，
.	。
?	？
!	！
:	：
;	；
"	" or "
'	' or '
( )	（ ）
[ ]	【 】
-	—
...	……

Usage

1. Check a Single File

Users can provide a file path or file content, and the skill will automatically analyze and generate a report.

2. Check Entire Project

Use the following command to check the entire project:

Check English punctuation in the entire project

The skill will automatically:

• Scan text files in the project (including .js, .ts script files)
• Exclude code files (.py, .java, .go, .cpp, .c, .cs, .php, .rb, .swift, .kt, .css, etc.)
• Exclude build directories (node_modules, .git, dist, target, build, out, etc.)

3. Scan Scope

Default file types to check (text files):

• .md - Markdown documents
• .txt - Text files
• .html, .htm - HTML files
• .xml - XML files
• .json - JSON files (only check string values)
• .yml, .yaml - YAML config files
• .properties - Properties files
• .vue, .jsx, .tsx, .js, .ts - Framework template files and script files

Default excluded directories:

• node_modules
• .git
• dist
• target
• build
• out
• .idea
• .vscode
• __pycache__
• .next
• .nuxt

Default excluded file types (code files):

• .css, .scss, .less, .sass
• .py
• .java
• .go
• .cpp, .c, .h, .hpp
• .cs
• .php
• .rb
• .swift
• .kt
• .rs
• .sql

4. Generate Report

After checking, generate a Markdown format report including:

• Original content of incorrect punctuation
• Location info (file path, line number, column number)
• Context snippet
• Suggested correct Chinese punctuation replacement

5. Batch Fix

Provide one-click fix functionality to automatically replace all detected incorrect punctuation with correct Chinese punctuation.

Output Format Example

# Chinese Punctuation Check Report

## Summary
- Files checked: 5 files
- Total errors found: 15
- Check time: 2026-01-01 10:00:00

---

## Error Details

### File: docs/README.md

| Line | Column | Wrong | Correct | Context |
|-----|-----|---------|---------|-------|
| 5   | 12  | ,       | ，      | "Hello, world" |
| 10  | 8   | .       | 。      | "This is a sentence" |

### File: content/about.txt

| Line | Column | Wrong | Correct | Context |
|-----|-----|---------|---------|-------|
| 3   | 15  | !       | ！      | "Welcome!" |

---

## Batch Fix

15 incorrect punctuation marks detected. Do you want to fix all? (yes/no)

Execution Flow

1. Scan Project: Traverse project directory, filter text files
2. Filter Exclusions: Exclude code files and specified directories
3. Read Content: Read each qualified file
4. Detect Errors: Use regex to identify English punctuation marks
5. Record Location: Record file path, line number, column number for each error
6. Generate Report: Output Markdown format inspection report
7. Provide Fix: Offer batch fix option

Smart Detection Rules

To avoid false positives, the skill uses the following smart detection rules:

1. Only check lines containing Chinese characters: Skip lines without Chinese characters
2. Exclude URLs and file paths: Do not replace punctuation in URLs or file paths
3. Exclude strings in code: Skip code strings even in allowed file extensions
4. Exclude comments: Skip Markdown code blocks, HTML comments, etc.