REST endpoint to get user permissions?

Konrad Garus July 17, 2017

How can my server for a Connect add-on determine what permissions a user has? I'm looking for ADMINISTER specifically, but it's unlikely to be right-specific.

I found two options, none of which seems ideal:

  • /rest/api/2/user/permission/search - according to docs it requires "admin" to access, which is way too strong a scope to ask for an add-on that claims to never ever even write to JIRA.
  • /rest/api/2/mypermissions - requires ACT_AS_USER, which seems pretty strong as well (permission to perform any action on behalf of any user? ouch!).

Am I missing something? Are these the only options?

1 answer

0 votes
Stephen Deutsch
Rising Star
Rising Star
Rising Stars are recognized for providing high-quality answers to other users. Rising Stars receive a certificate of achievement and are on the path to becoming Community Leaders.
July 17, 2017

I would recommend skipping the REST API completely, and use conditions to determine that:

https://developer.atlassian.com/static/connect/docs/latest/concepts/conditions.html

What you can do is to have 2 different entry points into whatever webItem/webPanel you are providing, and either separate it into different files (if you are referring to html files) or use the same base URL and include different query parameters in each one. One would check whether the user is admin, and the other one would check the same condition, but with invert:true.

So for example, the one with admin permission would point to:

/path-to-my-addon?isAdmin=true&issue=...

and the other inverted condition would point to:

/path-to-my-addon?isAdmin=false&issue=...

and then you can just read it out of the query parameters.

Stephen Deutsch
Rising Star
Rising Star
Rising Stars are recognized for providing high-quality answers to other users. Rising Stars receive a certificate of achievement and are on the path to becoming Community Leaders.
July 17, 2017

Of course, this only works if you are doing something that interacts with the user in the browser and is going to present the information right away. If it's reacting to a webhook or something like that, then you probably wouldn't be able to use what I mentioned. In that case, a little bit more information about your use case would be required.

Konrad Garus July 17, 2017

Thanks, but this is pretty weak security. You can't really trust requests coming from the client, assuming that they have always been sent by your app, properly validated on the client side, etc.

Stephen Deutsch
Rising Star
Rising Star
Rising Stars are recognized for providing high-quality answers to other users. Rising Stars receive a certificate of achievement and are on the path to becoming Community Leaders.
July 17, 2017

I do see your point. If you are using JWT authentication for the requests, though, then you can validate that the requests haven't been tampered with, I think.

Perhaps there is another way to get this information; I'll take another look.

Suggest an answer

Log in or Sign up to answer