CrossSpeak
Unicode homoglyph encoder/decoder (ASCII digits preserved; configurable < > encoding).
1:1 reversible • digits preserved • local HTML
Plain Text
Encode →
Swap
Clear
Ctrl
+
E
encode
Ctrl
+
D
decode
Ctrl
+
S
swap
Chars:
0
CrossSpeak Text
← Decode
Copy
Clear
Detection:
none
•
Digits:
ASCII preserved
•
<> mode:
untouched
Chars:
0
Angle bracket mapping
(
<
and
>
)
Untouched
< >
Option A
⟨ ⟩
Option B
‹ ›
Option C
＜ ＞
Why this matters:
models often treat
<
/
>
as structural tokens (HTML, comparisons, generics). Keeping them untouched is safest for code; Options A–C can be useful for “stealth text” while still staying LLM-readable.