nav[aria-label="Primary Navigation"] { padding: 0; & ul { list-style: none; width: 100%; display: flex; flex-direction: row; justify-content: start; align-items: start; gap: 30px; padding: 0; & li { margin: 0; } & ul li { list-style: none; } } }

Dark corners of Jira Automation

Mykenna Cepek · 2026-10-12T18:57:21+00:00

* I wanted to share a recent Jira Automation moment. I've been using AI on-and-off for a few days to help perfect a Jira automation flow. Despite my years of consistent and deep experience, all the inconsistencies consistently frustrate! Even my AI buddy sensed it, once I finally solved the last problem: I love that AI…

Mykenna Cepek

I wanted to share a recent Jira Automation moment. I've been using AI on-and-off for a few days to help perfect a Jira automation flow. Despite my years of consistent and deep experience, all the inconsistencies consistently frustrate! Even my AI buddy sensed it, once I finally solved the last problem:

I love that AI used that particular emoji.

The final fix resulted in splitting one Set Variable into two:

Set Variable FOO to something.
Set Variable FOO to FOO.replaceAll(…)

It did not work to do this in a single Set Variable step. Sigh.

To make this a discussion, I invite you to share a Jira automation "dark corner" you recently encountered, including the fix or workaround you ended up using.

P.S. I was surprised that there was no Discussion Group for Jira Automation. Is that a thing that would interest folks here? There is an Automation Q&A group, but that's different than a group for automation SMEs.

Comments

Becker, Rene

To be honest, I have abondoned Jira/CFL Automations more or less.
For very, very, very simple Flows, I still use it, but for anything longer than a two step automation, I forward everything to my own automation system. I am even in the process, to migrate existing Flows.

When the new automation billing hits, I'll migrate anyway because Jira Flows do force me to use way more steps than neccessary (for example the double assignments for variables, logging, looping, …) and there are usually more corrective steps than actual execution steps. The whole process of debugging in Flows is terrible.

Don't get me wrong, I understand that native Flows are made in this way, to allow non-technical staff to use it but it does only suffice for simplicity.

You asked for dark corners:

Flows do not always run. There is no apparent reason and there is no way to find out why, when there was neither success nor failure (or maybe I don't know how).
Using loops is like guesswork because the smartvalues do not always work the way, that I'd suggest they should.

Amanda Stroud

I'll tell you the weirdest one I have figured out. You see, branches are great until you want to iterate over work items in order, in which case this isn't really possible to do. In my use case, I was trying to figure out how to retain linked relationships in cloned work items. I basically created two fields that stored the new work items in a comma separated list on the original work item. One storing the new key's of the newly made linked work items, the other one storing the relationship type. There is a 3rd field that also stores the main work item that the other work items are going to be linked to.

So my problem came when I have an automation that triggers over the original work item. I want to do a JQL branch over the list of the newly linked work items, but the only way to correlate to the list of the link type is by looping in order of the index. And unfortunately, indexOf does not work to grab the index of an item in a list. But then I noticed that indexOf returns a list of -1's and then randomly has a 1 or a 0 in the list. This 1 or 0 is the location of where the work item is in the list I had stored. And you know what indexOf does work on? Strings! So I converted the -1 into a 5 (randomly chosen since 0 or 1 both indicate the location of the item in the list) and then I also have to have a condition to check if it's a 0 or a 1 and return the string that way. So then when I do an indexOf on this string, I get the actual index location of where the current work item is on the list! So now I can use get() to grab the linked relationship type on the right index. And if you're curious, this is what my variable looks like. No joke. "isZero" and "isOne" is separate variables (basically, these variables return true or false if 0 or 1 exists) helps make it look "less" messy.

So if you ever need to know how to grab corresponding values in separate lists, here you go! Use wisely.

Julio Rosa

I actually make quite a lot of use of Jira automations; while they can sometimes fall a bit short—in which case I rely on n8n—there is still a great deal you can accomplish without needing third-party apps.

Welcome!

It looks like you're new here. Sign in or register to get started.

Quick Links

Find more posts tagged with

cloud

automation