BiDi Controls
Directional formatting characters that can invisibly reorder text. Used in "Trojan Source" attacks to make malicious code appear benign. Your text contained these control characters which have been removed.
Mixed Scripts
Text mixing multiple writing systems (e.g., Latin + Cyrillic). Often used for phishing by spoofing domains like "аpple.com" (using Cyrillic 'а'). This advisory means your text may contain confusable characters.
Default Ignorables
Invisible Unicode characters like zero-width spaces, soft hyphens, and format controls. These can hide watermarks, bypass text filters, or carry steganographic data.
TAG Characters
Special invisible characters (U+E0000-U+E007F) historically used for steganography and prompt injection attacks. Most legitimate use cases (like emoji flags) are preserved in safe mode.
Orphan Combining Marks
Diacritical marks without base characters, or excessive "Zalgo text" that causes rendering issues. These have been cleaned while preserving legitimate accented characters.
Confusables
Visually similar characters from different scripts that could be used for spoofing or confusion attacks. The Spoofchecker has flagged your text as potentially suspicious.
Noncharacters
Unicode code points reserved for internal use (U+FDD0-U+FDEF and U+xxFFFE/U+xxFFFF). These should never appear in interchange text.
Private Use Area
Characters without standardized meaning (U+E000-U+F8FF and higher planes). In strict mode, these are removed for maximum security.
Non-ASCII Digits
Digits from non-Latin scripts (Arabic-Indic, Devanagari, etc.) that have been normalized to ASCII 0-9 in aggressive/strict modes.