Forum Discussion
Can Microsoft Frontier Program Copilot Cowork Agent Delegate Tasks to Other Copilot Agents?
Hi everyone,
I'm currently exploring the capabilities of Copilot Cowork that is available through the Microsoft Frontier Program, and I'm trying to understand whether a multi-agent orchestration pattern is officially supported.
My Use Case
I want users to interact with only a single, central Copilot Cowork agent.
For example:
- User asks the Cowork agent to create or update a Jira ticket.
- Instead of Cowork handling the Jira operation directly, it delegates or hands off the task to a dedicated Jira Copilot Agent.
- The Jira agent performs the required actions and returns the result.
- The Cowork agent then presents the final response back to the user.
Similarly, I would like to have specialized agents for:
- Jira
- ServiceNow
- Knowledge Management
- HR Operations
- Internal IT Support
- Other business systems
The goal is to have Cowork act as an intelligent orchestrator/router while specialized agents handle domain-specific operations.
Questions
- Is agent-to-agent delegation or handoff officially supported in Copilot Cowork (Frontier Program)?
- Can Cowork directly invoke another Copilot Studio agent?
- Is there any built-in multi-agent orchestration framework available today?
- If this is supported, what is the recommended architecture and implementation process?
- If it is not currently supported, what workarounds are people using?
- Power Automate?
- Agent as a tool/action?
- Custom APIs?
- Azure AI Foundry / Azure AI Agent Service?
- Other approaches?
I'm specifically looking for guidance from anyone who has worked with Copilot Cowork in the Frontier Program, since the documentation and public examples seem to focus mostly on standalone agents.
Any insights, architecture diagrams, documentation links, or real-world experiences would be greatly appreciated.
Thanks!
1 Reply
Hi,
Yes. In the Frontier Program, Copilot Cowork can call other Copilot Studio agents using the Connected Agent pattern or the Agent2Agent protocol. This allows Cowork to act as an orchestrator while specialized agents handle domain-specific tasks such as Jira and ServiceNow.
Multi-agent orchestration is a first-class pattern in Copilot Studio. You can implement inline (child) agents for tightly coupled subroutines and connected agents for separate, domain-specific capabilities that the orchestrator invokes.
The Agent2Agent (A2A) protocol enables cross-platform delegation and authenticated communication between agents. This is useful when integrating Azure AI Foundry or external agent platforms.
Step-by-step implementation:
- Design Cowork as the orchestrator and define routing rules and domain boundaries.
- Build domain-specific agents for Jira, ServiceNow, HR, and other systems. Publish them as connected agents.
- Connect agents using Agent2Agent or connected-agent links. Configure authentication and parameter passing while limiting context to only the required information.
- Integrate backend systems using Power Automate connectors for low-code scenarios or secure custom APIs for more complex operations.
- Add governance controls, including consent prompts for sensitive actions, role-based access checks, and centralized logging for auditing.
A practical starting point is a Cowork-to-Jira scenario. Implement the Jira agent using a Power Automate connector and connected-agent configuration, add logging and consent controls, and then expand to additional systems such as ServiceNow or HR. For more advanced orchestration requirements, Power Automate and Azure AI Foundry are commonly used to complement the solution.
Limitations:
- Preview feature: Multi-agent orchestration is currently available through the Frontier Program and is not yet generally available. Capabilities may change over time.
- Governance requirements: Least-privilege access, consent prompts, and logging must be configured as part of the solution design.
- Context passing: Cowork does not automatically transfer the entire conversation history to another agent. Required parameters and context must be explicitly defined.
- Complex workflows: Multi-step processes that span multiple systems may still require Power Automate or Azure AI Foundry for orchestration.
I hope this helps. Please let me know if you need any additional assistance.
If this reply helped resolve your question, please mark it as a solution. This helps other community members find answers more quickly.
Thank you.
Sunita