docs(agentic-orchestration): add monitor agents#7983
Conversation
afgambin
added
agentic orchestration
component:docs
|
👋 🤖 🤔 Hello, @afgambin! Did you make your changes in all the right places? These files were changed only in docs/. You might want to duplicate these changes in versioned_docs/version-8.8/.
You may have done this intentionally, but we wanted to point it out in case you didn't. You can read more about the versioning within our docs in our documentation guidelines. |
KunalNagi
left a comment
KunalNagi
left a comment
There was a problem hiding this comment.
Add screenshots for:
Execution tree with inner instances (Step 3)
toolCallResult JSON modal (Step 4)
agentContext JSON view (Step 5)
Visual context will significantly improve usability for non-technical operators
| 6. Select the **Are you satisfied with the result?** checkbox when you want to finish the process, then click **Complete task**. | ||
| 7. Go back to Operate. You will see the process instance is now completed, and the end event has been triggered. | ||
|
|
||
| ## Next steps |
There was a problem hiding this comment.
Suggestion: Right before Next Steps, based on enterprise production use cases, can we add a short section titled:
Production Monitoring Recommendations
For example:
Monitor escalation paths and repeated retries.
nspect evaluation outcomes (if using judge agents).
There was a problem hiding this comment.
SGTM, but I’d need more input on this. Also, regarding judge agents: they aren’t included in the example presented here, so I would keep this for another guide, or cover it when we provide it as part of a model blueprint that we can use as an example.
| - Tool call inputs and results. | ||
| - Additional metadata, such as reasoning traces and token usage. | ||
|
|
||
| ## Step 6: Understand how agent memory is stored |
There was a problem hiding this comment.
Suggestion: Can we explicitly clarify the trade-off between visibility and scalability?
“In Process” → better Operate visibility, limited scalability for long conversations.
“Document Storage” → scalable for production but content is not rendered inline in Operate.
There was a problem hiding this comment.
This was added as a note box at the end of the Step 6
docs/components/agentic-orchestration/monitor-agents-operate.md
Outdated
Show resolved
Hide resolved
|
@KunalNagi Thanks for your input! I addressed your comments. I didn’t add any screenshots, since we try to avoid them for docs maintainability. Instead, I added the requested outputs as JSON snippets. Ready for another pass! |
|
The preview environment relating to the commit af85483 has successfully been deployed. You can access it at https://preview.docs.camunda.cloud/pr-7983/ |
2 participants
Description
This PR adds a new guide explaining how to monitor and debug your AI agent process instances using Operate. This is part of this epic.
The guide has been written using an example running within the Embedded Intelligence organization environment, and a cluster version 8.9-alpha3.
When should this change go live?
bugorsupportlabel)available & undocumentedlabel)holdlabel)low priolabel)PR Checklist
{type}(scope): {description}commit message(s)/docsdirectory (version 8.9)./versioned_docsdirectory.@camunda/tech-writersunless working with an embedded writer.