Regression: Rovo agent (New Model)

Question

Regression: Rovo agent (New Model) no longer lists all Jira attachments

I’m currently facing an issue with the new backend model used by Rovo agents (rolled out around April 14, 2026). In my use case, I have a custom agent that formats Jira bugs and is expected to retrieve and list all files from the Attachments metadata.

With the previous model, the agent was able to enumerate all attachments reliably even with relatively minimal instructions. However, with the new model, it is no longer able to retrieve files that are present in the primary Attachments field. It only seems to access attachments that are referenced through comments or activity, which leads to incomplete evidence listing.

I’ve spent a significant amount of time experimenting with prompt refinements throughout the day — including:

* enforcing strict enumeration rules
* adding validation steps
* specifying retrieval from both Attachments and Activity metadata
* prohibiting summarization or grouping

Despite these refinements, the new model still omits several attachments or replaces them with generalized entries, even when the files are clearly present in the ticket metadata.

In contrast, the older model handled this scenario much more reliably and produced complete attachment listings with far simpler instructions.

Is this a known regression or a change in how attachment metadata is exposed to the model?

Has anyone else observed reduced attachment metadata retrieval with the new model?
Is this a known limitation or change in how attachment data is exposed to agents? Any guidance or recommended workaround would be appreciated.

Dr Valeri Colon _Connect Centric_ · Answer

@Arkadiusz Wroblewski I agree—this really comes down to expectations. Rovo is powerful, but it’s still context-driven, not deterministic. If you need “always list every attachment exactly,” that’s more of a system/process requirement than something AI can reliably guarantee today. It can look stable until a model shift changes behavior, so I’d treat this as a design boundary and use API/automation for consistency.

Arkadiusz Wroblewski · Answer

@Dr Valeri Colon _Connect Centric_

That’s actually my main concerns. Models are already quite powerful, but in my view they still lack the reliability needed to support stable processes.

Even at the most basic level, they are changing so quickly that there is not enough stability or predictability. What is needed here is a process, not interpretation, and certainly not something presented as a stable process when the outcome is not consistently reliable.

Dr Valeri Colon _Connect Centric_ · Answer

@sankaracourtallam_rathinasabapathy welcome. And I agree with @Arkadiusz Wroblewski I’d be careful calling this a confirmed regression. Rovo’s handling of Jira attachments isn’t fully consistent today, especially when relying on model interpretation rather than explicit retrieval. What you’re seeing (attachments in comments vs. primary field) has come up before and often points to how data is exposed to the agent, not just prompting. If you need full reliability, this is better handled via API/action. If reproducible, raise with support.