Did anyone set this up with a jetbrains product yet? What were your experiences?

[Darkshadenl]

I tried setting it up myself with JetBrains Rider. I have to say that JetBrains' AI integration is definitely better than it used to be, but it's still not very good. For example, in the skills screen you can uncheck skills, but those checkboxes often get automatically re-enabled again. So if you have a lot of skills and you just unchecked them, they turn themselves back on, which I find really annoying.

Regarding this topic though, I did manage to get it working. I added the MCP server. However, Codex (through the ai assistant) isn't smart enough to figure out by itself when it should use something, even though it has those skills. Naturally, I'd expect those skills to be sufficient for it to use memory. But maybe it simply has too many skills because of the issue mentioned above.

Are there other people who have set this up properly? Where it actually works well and just functions reliably?