Jira Automation copy Comments

Hello @Federico Frade , you are one step from done, and the step has two versions depending on how exact you need the result. Credit where due: @Gor Greyan branch is the right iteration and @Lalithesh smart value is the right source, the remaining problem is only where the comments land, so let me close that gap.

Why your Lookup attempt failed first, because it explains both fixes: smart values like the lookup result do not reliably survive into an advanced branch's iterations. The reliable carrier is a variable created before the branch, variables persist everywhere in the rule, including inside branches.

Option 1, no web requests, and it matches what you actually asked for (author names included): post ONE comment on the destination containing every source comment, attributed and chronological, using smart-value list iteration instead of a branch:

1. In the rule, after your destination work item exists (your clone or lookup step), add a branch: For most recently created work item (or your destination branch)
2. Inside it, a single Comment on work item action with:

Comments copied from {{triggerIssue.key}}:

{{#triggerIssue.comments}}
{{author.displayName}} ({{created.format("dd MMM yyyy HH:mm")}}):
{{body}}
----
{{/}}

The {{#list}}...{{/}} iteration renders every comment with its author and timestamp. One comment on the destination, full attribution, zero API calls. This is the version I would ship for a migration flow.

Option 2, if it must be one destination comment per source comment: the wall you hit with ADF disappears when you use the v2 REST endpoint, which accepts plain text bodies:

1. Create variable before the branch: name destKey, value your destination key (from your lookup or created item)
2. Advanced branch over {{triggerIssue.comments}}, variable name comment
3. Inside the branch, Send web request: POST to /rest/api/2/issue/{{destKey}}/comment, JSON body:

{ "body": "Comment by {{comment.author.displayName}} on {{comment.created}}:\n{{comment.body.jsonEncode}}" }

(Authenticated with an API token via Basic auth headers.) Two caveats from real use: jsonEncode is not optional, the first comment containing a quote character will break the request without it, and advanced branch iterations are not guaranteed to run in order, which is why the timestamp prefix matters, the destination reads chronologically even if creation order shuffles.

Between the two: Option 1 is simpler, needs no credentials, and satisfies the attribution requirement you stated. Option 2 buys per-comment granularity at the cost of a web request and token management. Most teams that start with 2 end up at 1 🙂

Hi @Federico Frade

@Arkadiusz Wroblewski KB is the right direction here.

The important point is that Jira Automation can copy comments from one work item to another without using the REST API. Atlassian’s documented approach is to iterate through the source work item’s comments and add them individually to the destination work item.

In your rule, because the original work item is the one that triggered the automation, you can reference it with {{triggerIssue}} even while you are inside the branch for the destination Service Desk work item. Atlassian confirms that triggerIssue continues to point to the original triggering work item inside branches, and all normal issue properties are available on it.

So I would use an Advanced branch over:

{{triggerIssue.comments}}

Give the branch variable a name such as comment, then inside that branch add a Comment on work item action with something like:

Comment by {{comment.author.displayName}} on {{comment.created}}

{{comment.body}}

The Comment on work item action supports smart values, so there is no need to fetch the comments via REST or convert ADF manually.

Hi @Federico Frade 

Welcome to the Atlassian Community!

The Comments actions are separate from the Edit Work Item action. To update fields such as Description and Priority, use the Edit Work Item action. To copy comments, add a separate Add Comment to Work Item action

Cheers!