Extract @mentions from text

Extract mentions finds every @handle in pasted text and lists them one per line. The match rule is @[\w]+: a literal at-sign followed by one or more word characters (letters, digits, underscore). Mentions stop at the first space, punctuation or symbol that is not a word character. The transform runs in your browser.

Input
Line 1:1 LF cloud_done Saved locally
Result Extract @Mentions
0 lines 0 chars

How mention matching works here

The pattern @[\w]+ captures an at-sign followed by one or more characters in [A-Za-z0-9_]. Handles stop at the first space, punctuation, emoji or accented character. So @alice, @bob_dev and @news_2026 all match cleanly.

Email addresses can also be captured as mentions because [email protected] contains @example. If you need to separate emails from mentions, run extract emails first to remove email-shaped tokens, or use extract regex matches with a pattern that requires a word boundary before the at-sign, like (?:^|\s)@\w+.

Output is one handle per line in the order they appear, including the leading at-sign. Duplicates are kept; pipe the result through remove duplicate lines for a unique handle list.

How to use extract @mentions from text

  1. 1Paste the post, message or transcript into the input panel.
  2. 2The output panel shows every @ handle, one per line.
  3. 3Click Copy to copy the list.
  4. 4Click Download to save it as a plain-text file.
  5. 5For unique handles only, send the result to remove duplicate lines.

Keyboard shortcuts

Drive TextResult without touching the mouse.

Shortcut Action
Ctrl FOpen the find & replace panel inside the input Plus
Ctrl ZUndo the last input change
Ctrl Shift ZRedo
Ctrl Shift EnterToggle fullscreen focus on the editor Plus
EscClose find & replace, or exit fullscreen
Ctrl KOpen the command palette to jump to any tool Plus
Ctrl SSave current workflow draft Plus
Ctrl PRun a saved workflow Plus

What counts as a mention here

At-sign followed by word characters

Letters A-Z a-z, digits 0-9 and underscore are accepted. @alice, @news_2026 and @a1 all match.

Stops at non-word characters

Spaces, punctuation, emoji and accented letters terminate the match. @josé is captured as @jos; @bob. is captured as @bob.

Email local-parts also match

[email protected] contains @example, which matches the pattern. To exclude emails, run extract emails first to find them, or remove them with find and replace.

Bare @ does not match

A lone at-sign with no following word characters is dropped. "don't @ me" yields no match for that @.

Order preserved, duplicates kept

Mentions appear in source order. Duplicates are not removed; for a unique list pipe the output into remove duplicate lines.

Worked example

The mention inside an email address (@example from [email protected]) is captured. The bare @ on its own is dropped because the pattern needs at least one word character after.

Input
Big thanks to @alice and @bob_dev.
Follow @news_2026 for updates.
Mail goes to [email protected] (and @ me later).
Output
@alice
@bob_dev
@news_2026
@example

Settings reference

Behaviour Effect on output
Handle body characters Letters A-Z a-z, digits 0-9, underscore.
Accented letters Not in the word class. @josé matches as @jos.
Punctuation and emoji Terminate the match.
Email local-parts Captured as mentions. [email protected] yields @example.
Bare @ Not matched.
Order and duplicates Source order kept, duplicates kept.

FAQ

How do I exclude email addresses from the mention list?
Two options. Either run extract emails first, then use find and replace to remove every captured email from the source before extracting mentions. Or switch to extract regex matches with the pattern (?:^|\s)@\w+, which requires whitespace or line start before the @.
Why does @josé get clipped?
The match rule uses the ASCII word class [\w]. Accented letters and other Unicode letters terminate the handle. For full Unicode handles, use extract regex matches with @[\p{L}\p{N}_]+ and the gu flag.
Is the leading at-sign included in each match?
Yes. Each line starts with @. Strip them with find and replace if you want the bare handles.
Are duplicates removed?
No. Every handle is kept in source order. Use remove duplicate lines for a unique list.
Is anything sent to a server?
No. The match runs entirely in your browser.