Merge pull request #5936 from MicrosoftDocs/main

Auto Publish – main to live - 2025-07-09 05:05 UTC

Author: PhilKang0704

.openpublishing.redirection.json

@@ -460,6 +460,11 @@
       "redirect_url": "../foundry-models/overview",
       "redirect_document_id": false
     },
+    {
+      "source_path": "articles/ai-foundry/concepts/concept-model-distillation.md",
+      "redirect_url": "../openai/how-to/stored-completions",
+      "redirect_document_id": false
+    },
     {
       "source_path": "articles/ai-foundry/model-inference/quotas-limits.md",
       "redirect_url": "../foundry-models/quotas-limits",
@@ -486,4 +491,4 @@
       "redirect_document_id": false
     }
   ]
-}
+}

articles/ai-foundry/agents/how-to/tools/model-context-protocol-samples.md

@@ -17,7 +17,7 @@ ms.custom: azure-ai-agents-code
 
 Use this article to find step-by-step instructions and code samples for connecting Foundry Agent service with MCP.
 
-Follow the [REST API Quickstart](../../quickstart.md?pivots=rest-api#api-call-information) to set the right values for the environment variables `AGENT_TOKEN`, `AZURE_AI_FOUNDRY_PROJECT_ENDPOINT` and `API_VERSION`.
+Follow the [REST API Quickstart](../../quickstart.md?pivots=rest-api#api-call-information) to set the right values for the environment variables `AGENT_TOKEN`, `AZURE_AI_FOUNDRY_PROJECT_ENDPOINT`, and `API_VERSION`.
 
 
 ## Create an Agent with the MCP tool enabled
@@ -35,7 +35,7 @@ curl --request POST \
               "type": "mcp",
               "server_label": "<unique name for your MCP server>",
               "server_url": "<your MCP server url>",
-              "require_approval": "never",
+              "allowed_tools": ["<tool_name>"], # optional
           }
       ],
   "name": "my-assistant",
@@ -68,7 +68,13 @@ curl --request POST \
 
 ## Create a run and check the output
 
-Create a run to pass headers for the tool and observe that the model uses the Grounding with Bing Search tool to provide a response to the user's question.
+Create a run to pass headers for the tool and observe that the model uses the Grounding with Bing Search tool to provide a response to the user's question. 
+`require_approval` parameter is optional. If not provided, `always` is the default value, meaning each time developer needs to approve before calling. Supported values:
+
+- `always` by default
+- `never` meaning no approval is required
+- `{"never":[<tool_name_1>, <tool_name_2>]}` you can also provide a list of tools without required approval
+- `{"always":[<tool_name_1>, <tool_name_2>]}` you can provide a list of tools with required approval
 
 ```bash
 curl --request POST \
@@ -81,9 +87,11 @@ curl --request POST \
           "mcp": [
             {
                 "server_label": "<the same unique name you provided during agent creation>",
+		"require_approval": "always" #always by default
                 "headers": {
                     "Authorization": "Bearer <token>",
                 }
+
             }
           ]
       },
@@ -97,6 +105,62 @@ curl --request GET \
   -H "Authorization: Bearer $AGENT_TOKEN"
 ```
 
+If the model is trying to invoke a tool in your MCP server with approval required, you get a run with `require_action` status.
+```bash
+{
+  "id": "run_123",
+  "object": "thread.run",
+  ...
+  "status": "requires_action",
+  ...
+  "required_action": {
+    "type": "submit_tool_approval",
+    "submit_tool_approval": {
+      "tool_calls": [
+        {
+          "id": "call_123",
+          "type": "mcp",
+          "arguments": "{...}",
+          "name": "<tool_name>",
+          "server_label": "<server_label_you_provided>"
+        }
+      ]
+    }
+  },
+  ...
+  "tools": [
+    {
+      "type": "mcp",
+      "server_label": "<server_label_you_provided>",
+      "server_url": "<server_url_you_provided>",
+      "allowed_tools": null
+    }
+  ],
+ ...
+}
+```
+Make sure you carefully reviewed the tool and argument(s) to be passed and make an informed decision for approval.
+
+## Submit your approval
+If you decide to approve, you need to set the `approve` parameter to be `true` with the `id` for the preceding tool calls.
+```bash
+curl --request POST \
+  --url $AZURE_AI_FOUNDRY_PROJECT_ENDPOINT/threads/thread_abc123/runs/run_abc123/submit_tool_outputs?api-version=$API_VERSION \
+  -H "Authorization: Bearer $AGENT_TOKEN" \
+  -H "Content-Type: application/json" \
+  -d '{
+	"tool_approvals": [
+        {
+            "tool_call_id": "call_abc123",
+            "approve": true,
+            "headers": {
+            }
+        }
+    ]
+}
+```
+
+
 ## Retrieve the agent response
 
 ```bash

articles/ai-foundry/agents/how-to/tools/model-context-protocol.md

@@ -27,7 +27,7 @@ You can bring multiple remote MCP servers to Foundry Agent service by adding the
 
 > [!Note]
 > * You need to bring a remote MCP server (an existing MCP server endpoint)
-> * With current MCP tool in Foundry Agent, explicit approval is not supported (only `never` is accepted for `require_approval` parameter). Please review carefully what MCP server(s) you added to Foundry Agent service. We recommend reviewing all data being shared with remote MCP servers and optionally logging it for auditing purposes.
+> * With current MCP tool in Foundry Agent, approval is required by default. Please review carefully what MCP server(s) you added to Foundry Agent service. We recommend reviewing all data being shared with remote MCP servers and optionally logging it for auditing purposes.
 > * Supported regions: `westus`, `westus2`, `uaenorth`, `southindia` and `switzerlandnorth`
 > * The MCP tool supports custom headers for a specific run, allowing you to pass headers as needed by MCP server, such as authentication schema. Headers you pass in will only be available for the current run and will not be persisted.
 
@@ -43,11 +43,20 @@ You can bring multiple remote MCP servers to Foundry Agent service by adding the
 1. Find the remote MCP server you want to connect to, such as GitHub MCP Server. Create or update a Foundry Agent with a `mcp` tool with the following information:
    1. `server_url`: the url of the MCP server, for example, `https://api.githubcopilot.com/mcp/`
    2. `server_label`: a unique identifier of this MCP server to the agent, for example, `github`
-   3. `require_approval`: only `never` is supported right now
+   3. `allowed_tools`: optional, a list of tools you want to allow without approval
 
 1. Create a run and pass additional information about the `mcp` tool in `tool_resources` with headers
    1. `tool_label`: use the identifier you provided during create/update agent
    2. `headers`: pass a set of headers required by the MCP server
+   3. `require_approval`: optional, if not provided, `always` is the default value, meaning each time developer needs to approve before calling. Supported values:
+      1. `always` by default
+      2. `never` meaning no approval is required
+      3. `{"never":[<tool_name_1>, <tool_name_2>]}` you can also provide a list of tools without required approval
+      4. `{"always":[<tool_name_1>, <tool_name_2>]}` you can provide a list of tools with required approval
+
+1. If the model is trying to invoke a tool in your MCP server with approval required, you will get Run status as `require_action`. Within `require_action` field, you can get more details on which tool in MCP server to be called, argument(s) to be passed and `call_id`. Make sur eyou review the tool, argument(s) and make an informed decision for approval. 
+
+1. Submit your approval to the agent with `call_id` by setting `approve` to `true.
 
 ## Next steps
 

articles/ai-foundry/concepts/concept-model-distillation.md

@@ -5,7 +5,7 @@ description: Learn how to use Azure OpenAI's advanced o3-mini, o1, & o1-mini rea
 manager: nitinme
 ms.service: azure-ai-openai
 ms.topic: include
-ms.date: 06/17/2025
+ms.date: 07/08/2025
 author: mrbullwinkle    
 ms.author: mbullwin
 ---
@@ -41,14 +41,14 @@ Azure OpenAI `o-series` models are designed to tackle reasoning and problem-solv
 
 | **Feature**  | **codex-mini**, **2025-05-16**  | **o3-pro**, **2025-06-10**   | **o4-mini**, **2025-04-16**  | **o3**, **2025-04-16** | **o3-mini**, **2025-01-31**  |**o1**, **2024-12-17**   | **o1-preview**, **2024-09-12**   | **o1-mini**, **2024-09-12**   |
 |:-------------------|:--------------------------:|:------:|:--------|:-----:|:-------:|:--------------------------:|:-------------------------------:|:---:|
-| **API Version** | `2025-04-01-preview` & [v1 preview](../api-version-lifecycle.md#api-evolution)   | `2025-04-01-preview`  & [v1 preview](../api-version-lifecycle.md#api-evolution)  | `2025-04-01-preview`   | `2025-04-01-preview`   | `2024-12-01-preview` or later <br> `2025-03-01-preview` (Recommended)   | `2024-12-01-preview` or later <br> `2025-03-01-preview` (Recommended) | `2024-09-01-preview` or later <br> `2025-03-01-preview` (Recommended)    | `2024-09-01-preview` or later <br> `2025-03-01-preview` (Recommended)    |
+| **API Version** | `2025-04-01-preview` & [v1 preview](../api-version-lifecycle.md#api-evolution)   | `2025-04-01-preview`  & [v1 preview](../api-version-lifecycle.md#api-evolution)  | `2025-04-01-preview` & [v1 preview](../api-version-lifecycle.md#api-evolution)   |  `2025-04-01-preview` & [v1 preview](../api-version-lifecycle.md#api-evolution)   |  `2025-04-01-preview` & [v1 preview](../api-version-lifecycle.md#api-evolution)   | `2025-04-01-preview` & [v1 preview](../api-version-lifecycle.md#api-evolution) | `2025-04-01-preview` & [v1 preview](../api-version-lifecycle.md#api-evolution)    | `2025-04-01-preview` & [v1 preview](../api-version-lifecycle.md#api-evolution)  |
 | **[Developer Messages](#developer-messages)** | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | - | - |
 | **[Structured Outputs](./structured-outputs.md)** | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ | - | - |
 | **[Context Window](../concepts/models.md#o-series-models)** |  Input: 200,000 <br> Output: 100,000 | Input: 200,000 <br> Output: 100,000 | Input: 200,000 <br> Output: 100,000 | Input: 200,000 <br> Output: 100,000 | Input: 200,000 <br> Output: 100,000 | Input: 200,000 <br> Output: 100,000 | Input: 128,000  <br> Output: 32,768 | Input: 128,000  <br> Output: 65,536 |
 | **[Reasoning effort](#reasoning-effort)** | ✅| ✅| ✅| ✅ |✅ | ✅ | - | - |
 | **[Image input](./gpt-with-vision.md)** | ✅ | ✅ | ✅ | ✅ | - | ✅ | - | - |
 | Chat Completions API | - | - | ✅ | ✅ | ✅ | ✅ | ✅ | ✅ |
-| Responses API | ✅  | ✅  | ✅ | ✅  | - | - | - | - |
+| Responses API | ✅  | ✅  | ✅ | ✅  | ✅ | ✅ | - | - |
 | Functions/Tools | ✅ | ✅ |✅ | ✅ | ✅  | ✅  |  - | - |
 | Parallel Tool Calls | - | - | - | - | -  | -  |  - | - |
 | `max_completion_tokens` <sup>1</sup> |  ✅ | ✅ | ✅ | ✅ |✅ |✅ |✅ | ✅ |

articles/ai-foundry/openai/how-to/reasoning.md

@@ -5,7 +5,7 @@ description: Learn how to use Azure OpenAI's new stateful Responses API.
 author: mrbullwinkle
 ms.author: mbullwin
 manager: nitinme
-ms.date: 06/20/2025
+ms.date: 07/08/2025
 ms.service: azure-ai-openai
 ms.topic: include
 ms.custom:
@@ -51,6 +51,8 @@ The responses API is currently available in the following regions:
 - `gpt-4.1-nano` (Version: `2025-04-14`)
 - `gpt-4.1-mini` (Version: `2025-04-14`)
 - `gpt-image-1` (Version: `2025-04-15`)
+- `o1` (Version: `2024-12-17`)
+- `o3-mini` (Version: `2025-01-31`)
 - `o3` (Version: `2025-04-16`)
 - `o4-mini` (Version: `2025-04-16`)