Solution: A whole word

Here is a possible solution to the task.

Code

say 'this is fine' ~~ /<< is >>/;

🦋 You can find the source code in the file whole-word.raku.

｢is｣

The << and >> anchors require a word boundary on each side of is.
The letters is also sit inside this, but there they have no word boundary on the left, so the engine skips that occurrence and matches the standalone word is instead.
Both candidates print the same ｢is｣, so how do you know which one actually matched? Change the is hidden inside this — for example, turn the string into 'thus is fine'. The output is still ｢is｣, unchanged, proving that occurrence was never the match. Now change the standalone word instead (say, 'this it fine') and the match disappears (Nil). You can also ask the match where it starts: ('this is fine' ~~ /<< is >>/).from returns 5, the position of the standalone is, not 2.