Hi @Pooja Deshmane ,
Thanks for sharing the additional details. This actually helps narrow down the issue quite a bit.
Since the agent works when you click the Pace Summary Agent button and provide the record number, but fails when you type the request directly, I don't think the problem is with the agent execution itself.
It looks more like an agent discovery or orchestration issue in QA.
I would suggest checking a few things:
Agent proficiency: Compare the agent's proficiency description between DEV and QA. This is particularly important because the orchestrator uses this information to determine which agent should handle a user's request. Make sure the description clearly explains what the agent does and when it should be invoked.
AI Search: Verify that AI Search is enabled and configured correctly in QA. Agent discovery can depend on this functionality.
Agent instructions: Check whether the agent instructions clearly describe the requests it should handle. For example, the agent should be able to recognize something like "Pace summary for PACE12345" and associate it with the appropriate workflow.
Orchestrator logs: Try the same natural-language request in both DEV and QA, then compare the execution logs. This should help determine whether the orchestrator is failing to identify the agent or whether something else is preventing execution.
Deployment dependencies: Since you migrated using an Update Set, I would also double-check whether all the related agent metadata and configuration records were included. Some configuration dependencies may need to be reviewed separately.
There's also a useful ServiceNow article that specifically mentions the "No agents are available at the moment" error and points to agent proficiency, AI Search, prompts, and orchestrator limits as possible causes:
AI Agents FAQ and Troubleshooting – ServiceNow Community
My recommendation would be to start with the agent proficiency and AI Search configuration in QA. The fact that the agent executes successfully when manually selected is a strong indication that the issue is related to how the orchestrator discovers and selects the agent.
One more thing: have you tried using a more explicit request, such as "Use Pace Summary Agent to summarize PACE12345"? If that works, it would further suggest an issue with natural-language agent discovery.
Hope this helps! Please let us know what you find.
If this resolves your issue, consider marking the reply as Accepted Solution and Helpful, so others facing the same problem can benefit too.
Thanks!
