Custom recommendation link page for each alert threshold definition

I might not be understanding your request fully but....you might be able to do this now with the current features of LogicMonitor.  You can set a property on a device or group, you then can use it as a token in the alert message or webhook/integration.  For example, you can set a property of help_link = http://confluence.acme.com/whatever  then you can include ##help_link## in the alert or http webhook. You can create dozens of properties so if you have several links they could be help_link1 , help_link2, etc.

 @Mike Suding I can't access http://confluence.acme.com/whatever so I'll try elaborate one more time. Let's assume I only have 2 alert thresholds setting: 

Data source | Alert threshold (critical case) | Wiki Page Value

1. WinVolumeUsage | 90% | http://confluence.arthrex.io/WinVolumeUsagealert

2. TomcatLogDirSize | 2 GB | http://confluence.arthrex.io/How+to+Adjust+Log+Cleanup+Interval

I want to customize in LogicMonitor so that whenever each alert is generated, it'll include as a property "recommendation" field into the JSON message that is sent to PagerDuty. 

Can you educate me how can I customize this? (e.g. with a new screenshot).. the screenshot above shows a generic JSON message format that is sent to PagerDuty and I did not see any option to customize the recommendation field other than hard code it for now. 

For some unknown reason I have not been able to to attach a screenshot (sorry).  You can set the URL as a property on a device or group...the 'acme' thing was just a non-functioning example.  Then, in the JSON area of integration you would reference this property name and surround it with 2 #'s   ##help_url## or whatever matches the property you want to use. The key is that it matches.

Mike, I think Horace wants these to be different per instance and/or per datasource -- how would you in the current UI define those properties?  Perhaps via PropertySources, but then I don't think there is a way to export those as tokens to an integration.

I have gone a different way with my LMKB integration, which is based on Template::Toolkit (but could be any templating system).  I have a layout with various standard fragments -- the 'triage' fragment is selected via inherited device property reference, then the article used tokens to conditionally report the correct details.  For example, this is a scrubbed sample from an F5 pool check we have been working on:

[% IF lmtoken.datasource.match("Pool State-/Common/foo_") %]
> Typically, when this fails the backend servers are wedged due to Tomcat issues.
[% END %]

The overall output is considered as MarkDown and feeds HTML into our ticketing system.  I would much prefer it this was builtin to LM via Groovy, but I couldn't wait :).

Thanks Mark & Mike for your feedback!! (been working on another project & just caught up on Mark's update). I still think this is worthwhile to pursue LogicMonitor DEV team to implement this as every customer will need to have this custom wiki troubleshooting page configured for each specific alert threshold within a datasource. I'll bring this up to my technical account manager to discuss further.