API corrupted state?

Is it possible to get my API in a corrupted state? If so, I seem to have done it. I've had a number of scripts acting up, and nothing I've done has really fixed it. Things were working smoothly up until a week or two ago when I started getting all kinds of API errors that kill my sandbox. I've tried disabling every script but 1, and they still produce errors. I tried copying my game, and started adding the scripts again over there, and so far for the first couple I've tried they seem to work fine again in the new game. If there is a way to get my original game up and running again, I'd much prefer that to re-installing/configuring all the API scripts in the new copy I made. Any suggestions? Thanks!!

Can you list the errors you're getting?  What scripts are you using?

CombatMaster gives this error in the original game, seems to work fine in the copy. Edit : I should have mentioned that I get this error when I set a token marker on a token. I thought InspirationTracker was having problems, but after more experimentation it seems that if you try to use it on a character that never had inspiration it doesn't work. (I guess the inspiration attribute doesn't exist until the first time you hit the box on the 5e OGL sheet?) So I really think it's just CombatMaster w/ the issue.

Yes. It's possible if you switch maps while combat is active.  You switch maps with active combat and boom.  Simply restart API Sandbox and it will reset itself.   Best choice is to hold combat, then switch maps, then restart assuming all tokens are the same

I've run into this multiple times.  I've gotten used to the easy path of a restart, but I'll look at stopping the failure

I don't think changing tabs during combat is the issue. I'm not even in combat. I'm really just trying to get the token marker status changes to work, and that's what has been the #1 thing throwing errors. However, the behavior has changed. Right now adding a simple token marker ( not using custom markers) is not killing the sandbox, but is also not giving the status update in chat. And it's putting this error in the API sandbox now: "ERROR: You must include the query string in your URL. For example, use '<a href="https://s3.amazonaws.com/files.staging.d20.io/images/123456/med.png?12345678" rel="nofollow">https://s3.amazonaws.com/files.staging.d20.io/images/123456/med.png?12345678</a>', not '<a href="https://s3.amazonaws.com/files.staging.d20.io/images/123456/med.png" rel="nofollow">https://s3.amazonaws.com/files.staging.d20.io/images/123456/med.png</a>'"

@Erza, You realize that these two errors are beyond any API to manage?&nbsp; Just restart sandbox and all is good. I may not ever fix this.&nbsp; If I put something in that prevents failure, then I get posts why nothing is happening.&nbsp; Its a zero sum game

Victor B. said: @Erza, You realize that these two errors are beyond any API to manage?&nbsp; Just restart sandbox and all is good. I may not ever fix this.&nbsp; If I put something in that prevents failure, then I get posts why nothing is happening.&nbsp; Its a zero sum game There's no way this is true. If your script has a failure in it that kills the sandbox, it is possible to avoid that failure. It might not be easy to track down, and might require substantial rewrites (hard to know), but it can be done. At the very least, if you can prevent failure, you can have the script post that a failure has been presented. That would be far superior to killing the sandbox.

Restarting the sandbox doesn't help if trying to use CombatMaster in any way kills the sandbox again immediately. :) Regardless, the sandbox killing part seems to have gone away, fortunately (now idea how/why), just leaving me with the other error regarding image URLs. And I believe I know how I caused that part now. I had reversed the green/red turn images and evidently left off the query string, so I tried it again leaving on the query string and the other error has gone away. So things are stable for now, thanks!

Wait, you can't even restart the sandbox with CM?&nbsp; PM me a game link and I'll take a look.&nbsp; Something else is going on.&nbsp; Or if I'm reading this right, you changed the default token markers?&nbsp; I'm working on a fool proof solution for that issue because everyone has that issue if they change from default.&nbsp; Solution might be to not allow changes, just use them or shut them off, however, if I do come up with a custom solution I'll need to put in a verification process of some type.&nbsp; Not sure yet.&nbsp;&nbsp;

@Gig, CM has never crashed a sandbox on reboot.&nbsp; There's been some recent issues with Roll20 Github and one-click (now resolved).&nbsp; As far as custom token markers go, that's been a known issue.&nbsp; The problem is that Roll20 has a specific way of storing images and if that image address doesn't make it into CM perfectly then it fails, why if I even go forward with custom token markers, I need to put in a verification.&nbsp;&nbsp;