Hi, is there any way to orchestrate the created agents in StudioAssist? I have several agents but I have to switch manually between agents and using a file based context system for sharing information between agents wanted to know if there is a way to achieve this? I am using 11.4.0
Hi, thank you for the question. In Katalon Studio 11.4.0, Agent Profiles are designed to be selected manually for each chat session. StudioAssist does not currently provide built-in multi-agent orchestration, automatic profile routing, or direct handoffs between agents.
A practical workaround is to keep shared instructions and outputs in version-controlled project files, for example, a Markdown handoff/context file and instruct each profile to read and update it. This keeps the workflow repeatable and shareable across the team. An external MCP server can also give an agent access to systems such as Jira, GitHub, or a document store, but MCP does not itself orchestrate multiple Agent Profiles.
We appreciate the planning → test creation → review workflow you described; it is valuable feedback for future StudioAssist enhancements.
We would appreciate it if you could kindly create a post in our Feature Request - Katalon Community forum category with your preferred workflow and examples of the agent roles, handoff context, and expected outcome. This will help our product team assess the use case and allow other users to contribute their feedback.
Thank you
Bella
@manthanmevada can you request as Bella instructed?
Interesting use case. As of now, StudioAssist doesn’t appear to support true multi-agent orchestration.
I’d recommend submitting a Feature Request with your workflow details for Katalon product team review.
keep a shared handoff file (Markdown works well) in your project and instruct each agent profile to read and write to it. this makes the workflow repeatable and version-controllable
I tried another approach where I created a supervisor agent and fed it different skills with references and templates and it worked pretty well for me - thought it’s chat based it’s pretty impressive.
Are you guys planning to add a feature for KRE where we can invoke these agents in pipeline as well?
This is a really useful pattern, thanks for sharing! Would you be open to elaborating a bit more on how you’ve set this up in practice?
Better create a thread under tips and tricks and elaborate, so that it would be helpful for all
i agree. you can share how structured and workaround details for other’s knowledge