I assume you've already tried removing the memory limit and setting a large num_procs value? Sometimes you get inappropriate error messages when you exceed resource limits. But I admit this doesn't look like one of those errors.
You say MeCab works on the CLI of the Jobe server. Is it possible you have multiple versions of Python installed? I see that CodeRunner is trying to use python3.7 which is a fairly old version. Also, did you install mecab globally or just for the logged in user?
You say MeCab works on the CLI of the Jobe server. Is it possible you have multiple versions of Python installed? I see that CodeRunner is trying to use python3.7 which is a fairly old version. Also, did you install mecab globally or just for the logged in user?
You can check that MeCab works for a typical jobe worker thread with a command like
Make sure your code works from that python shell.
sudo -u jobe01 python3.7Make sure your code works from that python shell.