You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
# Build collaborative, multi-agent systems with Connected Agents
18
18
19
-
Connected agents in Azure AI Foundry Agent Service let you break down complex tasks into coordinated, specialized roles—without the need for a custom orchestrator or hand-coded routing logic. With this capability, you can design systems where a primary agent intelligently delegates to purpose-built sub-agents, streamlining workflows like customer support, market research, legal summarization, and financial analysis.
19
+
Connected agents in Azure AI Foundry Agent Service let you break down complex tasks into coordinated, specialized roles—without the need for a custom orchestrator or hand-coded routing logic. With this capability, you can design systems where a primary agent intelligently delegates to purpose-built subagents, streamlining workflows like customer support, market research, legal summarization, and financial analysis.
20
20
21
21
Rather than overloading one agent with too many skills, you can build focused, reusable agents that collaborate seamlessly—scaling both performance and maintainability.
22
22
@@ -74,7 +74,8 @@ Checks the contract against internal standards or uploaded guidelines to identif
74
74
## Limitations
75
75
76
76
* Connected agents cannot call local functions using the function calling tool. We recommend using the [OpenAPI tool](tools\openapi-spec.md) or [Azure Functions](tools\azure-functions.md) instead.
77
-
* It is currently not possible to guarantee citations will be passed from connected agents. You can try using prompt engineering combined with different models to try and improve the possibility that citations will be outputted by the main agent, but results are subject to variability.
77
+
* It is currently not possible to guarantee citations will be passed from connected agents. You can try using prompt engineering combined with different models to try and improve the possibility that citations will be outputted by the main agent, but results are subject to variability.
78
+
* Connected agents have a maximum depth of 2. A parent agent can have multiple subagent siblings, but subagents cannot have their own subagents. Exceeding this depth results in an `Assistant Tool Call Depth Error`.
78
79
79
80
:::zone pivot="portal"
80
81
@@ -83,7 +84,7 @@ Checks the contract against internal standards or uploaded guidelines to identif
83
84
84
85
1. Navigate to the **Agents** page in the portal
85
86
2. Select an existing agent from the list or create a new one.
86
-
3. Scroll down to the **Connected agents** section in the agent's setup panel and select **Add +**.
87
+
3. Scroll down to the **Connected agents** section in the agent's set up panel and select **Add +**.
87
88
88
89
:::image type="content" source="../media\connected-agents\connected-agents-foundry.png" alt-text="A screenshot of the agents page in the Azure AI Foundry." lightbox="../media\connected-agents\connected-agents-foundry.png":::
89
90
@@ -93,7 +94,7 @@ Checks the contract against internal standards or uploaded guidelines to identif
93
94
- Add a clear **description** of when and why the connected agent should be invoked. This helps guide the main agent’s decision-making on when to hand off tasks to connected agents during runtime.
94
95
5. Select **Add +**
95
96
6. Repeat steps 3–5 to add additional specialized agents to the main agent.
96
-
7. Once the connected agent(s) appear in the setup panel, scroll up and select **Try in Playground**
97
+
7. Once the connected agents appear in the setup panel, scroll up and select **Try in Playground**
97
98
8. Use test prompts in the Agent Playground to validate that the main agent correctly routes tasks to the connected agents when applicable. For example, if you’ve created a main agent called `research_agent`, which doesn't have any tools configured, and connected an agent named `stock_price_bot`, try a prompt like:
98
99
99
100
**"What is the current stock price of Microsoft?"**
Copy file name to clipboardExpand all lines: articles/ai-foundry/agents/how-to/tools/deep-research-samples.md
+2-1Lines changed: 2 additions & 1 deletion
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -574,7 +574,8 @@ with project_client:
574
574
:::zone-end
575
575
576
576
> [!NOTE]
577
-
> Limitation: The Deep Research tool is currently recommended only in nonstreaming scenarios. Using it with streaming can work, but it might occasionally time out and is therefore not recommended.
577
+
> * Limitation: The Deep Research tool is currently recommended only in nonstreaming scenarios. Using it with streaming can work, but it might occasionally time out and is therefore not recommended.
578
+
> * Currently, Azure AI Foundry Agent Playground UI only supports starting runs in streaming mode, and as a result, users might experience connection drops and timeouts. Use the supported methods above which uses runs in non-streaming mode.
0 commit comments