CI slow-test hypothesis fail: libcst fails to parse 𮯰

[jakkdl]

I got a failing test case in https://github.com/python-trio/flake8-async/actions/runs/13008484527/job/36280611992#step:5:68 and opened Instagram/LibCST#1288

Idk if/when libcst will get around to fixing it, versus how likely hypothesis is to re-generate examples with unicode code points that will break libcst, but given I only bumped CI to 3.13 3 days ago it might become a regular thing and warrant some logic in the test to xfail on weird unicode.

[Zac-HD]

ah, classic. I think this is "just" a new version of Unicode, unlike the new-parser issues I found in 3.9...