Invisible Character Inspector & Cleaner
Visualizes hidden zero-width spaces, directional marks, and control characters with inline high-contrast badges before you copy.
đź“– Engineering & Technical Deep-Dive Click to expand
How it works: Scans your text character by character using deep Unicode code point inspection to identify, visualize, and safely strip unprintable characters, Trojan Source bidirectional overrides, and hidden smuggling sequences. Detects Zero-Width Space (U+200B), Zero-Width Non-Joiner (U+200C), Zero-Width Joiner (U+200D), Byte Order Mark (U+FEFF), Word Joiner (U+2060), Combining Grapheme Joiner (U+034F), Soft Hyphens (U+00AD), Hangul Fillers (U+3164), Invisible Math Operators (U+2061–U+2064), Non-Breaking Spaces (U+00A0 and 14 Unicode space variants), Trojan Source bidi directional overrides (U+202A–U+202E, U+2066–U+2069), and prompt-injection tag characters (U+E0000–U+E007F). While neutralizing invisible threats, genuine emoji sequences (family emojis, skin tones, authentic subdivision flags) and international script marks (Hebrew/Arabic directional marks) remain completely preserved.
Why it matters: Invisible characters frequently slip in when copying text from Slack, Microsoft Teams, PDF documents, or rich-text web pages. In production systems, these hidden characters cause catastrophic yet baffling errors: database lookups fail silently because '[email protected]' with an invisible ZWSP doesn't match the database record; Python and YAML parsers crash on invisible indentation whitespace; and cryptographic signature verification fails due to hidden byte mismatches. Furthermore, cyber threats like Trojan Source exploit invisible bidirectional overrides to reorder source code visually while executing malicious logic, and prompt-injection attacks smuggle hidden instructions behind emojis using unprintable Unicode tag characters.
Security Architecture: In the separate Ad-Free Private Vault, zero external scripts or analytics are loaded. The entire inspection runs inside your browser sandbox.