lookupissues conditional check on Issue Type isn't working.

[LPK]

Hey All, 

I am working on a JIRA Automation + Power Automate in order to send the count of distinct issues worked on by my squad while closing the Sprint to the Teams in the form of a message on channel or group chat.

JIRA Automation rule sequence:

When: Sprint completed
Then: lookup issues
And: add value to audit log
And: Send Web Request

Add value to Log (Action):

Issue Type: {{fields.issuetype.name}} ; Stories: {{lookupIssues.where("fields.issuetype.name = 'Story'").size}}

Send Web Request: Webhook body (Custom Data)

{

"totalIssues": {{lookupIssues.size}},
"storyCount": {{lookupIssues.where("fields.issuetype.name = 'Story'").size}},
"bugCount": {{lookupIssues.where("fields.issuetype.name = 'Bug'").size}},
"taskCount": {{lookupIssues.where("fields.issuetype.name = 'Task'").size}},
"spikeCount": {{lookupIssues.where("fields.issuetype.name = 'Spike'").size}}
}

-----

So, when my rule is run, i see the the log value as ---> Issue Type: Story ; Stories : 0

Which means, the smart value to calculate the count of issues of type = Story is failing.

I have tried multiple ways with aid of ChatGPT & Copilot, but efforts have gone in vain. 

Note: {{issue.issueType.name}} didn't resolve to give me type = Story but {{fields.issuetype.name}} did.

Please help.

#JIRA #JiraAutomation #automation #powerautomate

[Akash Singh · Accepted solution]

@LPK Your use case would ideally require a reduce-like function that can iterate over the lookupIssues list, filter issues by issueType, and aggregate the counts. However, such advanced list operations are not currently supported in Jira Automation smart values.

As a workaround, I recommend breaking down your lookupIssues into separate actions using specific JQL queries for each issue type.

For example, your first Lookup Issues action can use:

project = XYZ AND sprint = {{sprint.id}} AND issuetype = "Story"

Then, add a Create variable action with:

• Name: storiesCount

• Value: {{lookupIssues.size}}

Repeat this process for other issue types like Bug, Task, and Spike, each with their own Lookup Issues and variable creation steps.

Once done, you can use these variables to construct your output:

"totalIssues": {{#=}}{{storiesCount}}+{{bugCount}}+{{taskCount}}+{{spikeCount}}{{/}},
"storyCount": {{storiesCount}},
"bugCount": {{bugCount}},
"taskCount": {{taskCount}},
"spikeCount": {{spikeCount}}

[LPK]

Amazing! @Akash Singh 
I just tried for Issue type = Story. It worked!

So, I will have to perhaps iterate this approach "lookupIssues -> Create a variable (to store the count) for all issue types and then send those smart values to the JSON payload.

I have been banging my head to solve this with an aid of ChatGPT or Copilot. 
But you've proved a human is superior to computers at any given point of time :) 

[LPK]

Finally, I achieved it. Once again, tons of thank you Mr @Akash Singh 

[Bill Sheboy]

Hi @LPK 

You may use smart value, list filtering to do this:

https://community.atlassian.com/forums/Automation-articles/Filtering-smart-value-lists/ba-p/1827588

The general approach uses conditional logic to filter by the issue type, and then a wrapped math expression to perform the count.  This will be faster than using multiple Lookup Work Items actions, with different JQL expressions, and reduce the chance of errors due to repeating the JQL multiple times.

For example:

• trigger: Sprint Completed
• action: lookup work items, with JQL to get the ones in the sprint from the trigger
• action: some action that needs the counts, such as...

Story count: {{#=}}0{{#lookupIssues}}{{#if(equals(issuetype.name,"Story"))}}+1{{/}}{{/}}{{/}}
Bug count: {{#=}}0{{#lookupIssues}}{{#if(equals(issuetype.name,"Bug"))}}+1{{/}}{{/}}{{/}}
...

How that works is...

• iterate over the lookup results
• using a condition to filter by the issue type's name equals what is needed
• and when found, +1 to the count
• wrapping that in a long-format, math expression, with a 0 at the front for a default value

And, as you are doing this in order to Send Web Request, I recommend pre-building the entire expression's JSON as a variable and then using that in the request.  This will help with both performance and debugging of the message.

Kind regards,
Bill