Preventing Character Encoding Mismatch (Mojibake)

This skill provides guidelines and procedures for avoiding character encoding issues (specifically Mojibake where UTF-8 Thai text is read/written as Windows-1252/ISO-8859-1).

Problem Description

Mojibake (e.g. Thai text becoming คุณต้องภาร) occurs when text encoded in UTF-8 is read or written using a single-byte encoding like Windows-1252 (CP1252 / ISO-8859-1 / Latin-1). On Windows systems, if file writing tools or terminal redirects do not explicitly specify the encoding, the OS defaults to ANSI/CP1252, corrupting multi-byte characters like Thai script.

Guidelines for Agents

1. File Writing & Overwriting

Whenever writing or editing files containing Thai (or any non-ASCII) text: