Inspection field guide sections 03-07

Hidden Unicode,
explained.

A practical reference for the cleaner: how each pass works, which characters it handles, where invisible Unicode comes from and what to review before copying.

03 / Method

Clean the text without flying blind.

The entire workflow stays on one inspection sheet, so the source, result and removal record remain easy to compare.

1
Paste or open
Paste a Claude response, drop a plain-text file, or load the built-in example. Word, line and character counts update immediately.
2
Choose a pressure
Careful removes common non-printing characters while preserving sensitive joiners. Thorough also strips joiners and variation selectors.
3
Review and copy
Inspect every affected code point and its positions, reveal the marks in context, then copy or download the cleaned result.

04 / Character ledger

Exactly what this Unicode cleaner removes.

Careful is the safer default. Thorough is intentionally destructive and should be reviewed when text contains emoji or joined scripts.

Unicode characters supported by the careful and thorough cleaning modes
Code point	Character	Careful	Thorough
`U+00AD`	Soft hyphen	Removed	Removed
`U+200B`	Zero-width space	Removed	Removed
`U+2060`	Word joiner	Removed	Removed
`U+FEFF`	Byte order mark	Removed	Removed
`U+200C`	Zero-width non-joiner	Preserved	Removed
`U+200D`	Zero-width joiner	Preserved	Removed
`U+FE00–FE0F` `U+E0100–E01EF`	Unicode variation selectors	Preserved	Removed

Also applied: Unicode NFC normalization standardizes canonically equivalent sequences. The report lists that separately; it is not counted as a removed hidden character.

05 / Context

Why hidden characters appear in copied text.

Non-printing Unicode is not automatically suspicious. Soft hyphens help with line breaks, byte order marks identify encoding, and joiners can shape emoji or languages such as Persian. Editors, PDFs, websites and copy-and-paste workflows may introduce them for ordinary reasons.

That is why the cleaner names each character and shows where it occurred instead of guessing why it is there.

06 / Limits

Why the Thorough pass deserves a review.

Some invisible Unicode carries real meaning. Joiners shape emoji and joined scripts, while variation selectors can control how a symbol is presented. Thorough removes those characters deliberately.

Careful leaves sensitive marks intact while removing the common ones. Whichever pressure you choose, compare the result and read the inspection report before copying.

07 / Desk questions

Questions about Claude text and hidden Unicode.

What happens if the text is already clean?

The cleaned copy stays the same, apart from equivalent Unicode sequences being standardized through NFC normalization. The report records that no supported hidden marks were found.

What is the difference between Careful and Thorough?

Careful preserves joiners and variation selectors that may carry meaning. Thorough removes those characters too, which can alter emoji presentation or text in joined scripts.

Why do ordinary documents contain hidden characters?

Browsers, PDFs, word processors and publishing systems use them for spacing, line breaks, encoding, emoji and joined scripts. Copy-and-paste can carry those marks into a new document.

Can I clean text from ChatGPT, Word or Google Docs?

Yes. The inspection is based on Unicode code points rather than the source application, so you can use it with plain text copied from any tool.

Does the cleaner rewrite my words?

No. It filters the selected non-printing characters and applies standard NFC normalization. It does not paraphrase, summarize or change vocabulary.

Why review the report before copying?

Some invisible characters are meaningful. The report makes the trade-off visible, and the marked-up preview shows each occurrence in its original context.