Teams

​

Introduction

• Discover Teams and Channels: List the teams the signed-in user is a member of, and list the channels within a team.
• Discover and Read Chats: List the user’s 1:1, group and meeting chats with member and last-message previews, and read messages within a specific chat.
• Search Across Teams: Search chat and channel messages across the user’s Teams activity using Microsoft’s KQL (Keyword Query Language) syntax.
• Send Chat Messages: Send a message to an existing chat, or to one or more people by email — automatically creating the 1:1 or group chat when needed.
• Post to Channels: Post a new message to a channel, or reply to an existing channel post, with optional @mentions.

​

Toolkit’s Account Setup and Configuration in Teams

​

Registering an App in Azure Active Directory (Azure AD)

1. Access Azure Portal: Open your web browser, navigate to the Azure Portal, and log in using an account with sufficient permissions to register applications in Azure AD.
2. Navigate to App Registrations: In the Azure portal, use the search bar at the top to search for “App registrations” and select “App registrations” from the search results under “Services”.
3. Create New Registration: On the “App registrations” page, click ”+ New registration”.
4. Configure App Registration Details: On the “Register an application” page, provide the following information:
   • Name: Enter a meaningful and descriptive name for your application registration, for example “ELITEA Teams Integration” or “ELITEA Agent Access to Teams”.
   • Supported account types: Select the appropriate account type based on your organization’s requirements. In most cases, “Accounts in this organizational directory only ([Your Organization Name] only - Single tenant)” is the recommended option for internal organizational use.
   • Redirect URI: Since Teams only supports Delegated (User OAuth) access, a redirect URI must always be configured:
     1. In the “Select a platform” dropdown, select “Web”.
     2. In the URI field, enter the ELITEA callback URL for your instance. The correct value depends on your deployment — refer to the info callout below.
5. Register Application: Click the “Register” button at the bottom of the page to create the app registration.

​

Application Credentials

• Application (client) ID — Copy and store this value securely.
• Directory (tenant) ID — Copy and store this value securely.

​

Configure API Permissions for the Registered App

1. Navigate to API Permissions: In your registered app within the Azure portal, navigate to the left-hand menu and click “API permissions”.
2. Add Permissions: On the “API permissions” page, click ”+ Add a permission”.
3. Select API Type - Microsoft Graph: In the “Request API permissions” panel, select the “Microsoft Graph” API tile.
4. Select Permission Type - Delegated permissions: Choose “Delegated permissions” — the Teams toolkit always acts on behalf of the signed-in user, never as the application itself.
5. Add Microsoft Graph Teams Scopes: Search for and select the scopes your Agents require, for example:

   Scope	Description	When to Grant
   User.Read / User.ReadBasic.All	Resolve people (recipients, mentions, senders) by email	Needed by every tool that accepts an email address (find_chat_messages, send_chat_message, send_channel_message, search_teams_messages)
   Team.ReadBasic.All	List the teams the signed-in user is a member of	Needed for list_teams
   Channel.ReadBasic.All	List the channels of a team	Needed for list_channels
   Chat.Read (or Chat.ReadWrite)	Read chats and chat messages	Needed for list_chats, find_chat_messages, search_teams_messages
   Chat.Create	Create a new 1:1 or group chat when one doesn’t already exist	Needed for send_chat_message when messaging people instead of an existing chat
   ChatMessage.Send	Send a chat message	Needed for send_chat_message
   ChannelMessage.Send	Post or reply in a channel	Needed for send_channel_message

   The exact scope names are entered into the Scopes field of the ELITEA credential — see Step 1 below. Grant only the scopes needed for the tools you plan to enable.
   Reading channel messages is not supportedThere is intentionally no scope listed above for reading channel messages. That capability requires ChannelMessage.Read.All, an application permission that needs tenant admin consent, and is out of scope for this toolkit.
6. Add Permissions: Click “Add permissions” at the bottom of the panel to add the selected scopes to your application registration.
7. Grant Admin Consent (if required by your tenant): On the “API permissions” page, click “Grant admin consent for [Your Organization Name]” and then “Yes” if your organization requires admin consent for delegated permissions before users can authorize the app.

​

Configure the Client Secret

1. Navigate to Certificates & secrets: In your registered app within the Azure portal, click “Certificates & secrets”.
2. Create New Client Secret: On the “Certificates & secrets” page, select the “Client secrets” tab and click ”+ New client secret”.
3. Configure Client Secret Details: In the “Add a client secret” panel, enter a Description and choose an appropriate Expiration period.
4. Add Client Secret: Click “Add” to create the client secret.
5. Securely Copy and Store the Client Secret Value: Immediately copy the generated Client Secret Value displayed in the “Value” column. This is the only time you will see the full value. Store it securely, preferably in ELITEA’s Secrets feature.

​

System Integration with ELITEA

​

Step 1: Create Teams Credentials

1. Navigate to Credentials Menu: Open the sidebar and select Credentials.
2. Create New Credential: Click the + Credential button.
3. Select Teams: Choose Teams as the credential type.
4. Configure Credential Details:

   Field	Required	Description
   Display Name	Required	Descriptive name for this credential (e.g., “Teams - Support Account”)
   ID	Required	Unique identifier for the credential. Auto-populated from the Display Name
   Client ID	Required	Azure AD Application (client) ID from your app registration
   Client Secret	Required	Azure AD Client Secret value generated in your app registration
   Auto Refresh Token	Optional (default: enabled)	Automatically refresh the access token using the offline_access scope
   OAuth Discovery Endpoint	Optional	OAuth discovery endpoint, in the format https://login.microsoftonline.com/{tenant_id}, where {tenant_id} is the Directory (tenant) ID from your app registration
   Scopes	Optional	OAuth scopes to request (e.g., User.Read, Chat.Read, ChatMessage.Send, Chat.Create, Team.ReadBasic.All, Channel.ReadBasic.All, ChannelMessage.Send)

5. Log in: Once the OAuth Discovery Endpoint field is filled in, a Login button appears next to the Test Connection button. Click Login to initiate the OAuth authorization flow — ELITEA calls the Teams check-connection endpoint, which redirects to Azure AD for user sign-in. After you authorize the application, the token is stored and the connection is verified against Microsoft Graph.
   • Test Connection: Click Test Connection to verify that your credentials are valid and ELITEA can successfully connect to Microsoft Teams. This performs a health check against GET /me/joinedTeams on Microsoft Graph.
6. Save Credential: Click Save to create the credential. It will be added to the credentials dashboard and available to use in toolkit configurations.

​

Step 2: Create Teams Toolkit

1. Navigate to Toolkits Menu: Open the sidebar and select Toolkits & Indexes.
2. Create New Toolkit: Click the + Toolkit button.
3. Select Teams: Choose Teams from the list of available toolkit types.
4. Configure Toolkit Details:

   Field	Required	Description
   Toolkit Name	Required	Descriptive name for the toolkit (e.g., “Teams - Support Account”)
   Description	Optional	Brief description of the toolkit’s purpose
   Credential	Required	Select the Teams credential created in Step 1

5. Enable Desired Tools: In the “Tools” section, select the checkboxes next to the specific Teams tools you want to enable. Enable only the tools your agents will actually use to follow the principle of least privilege.
   • Enable MCP access for selected tools — (optional) exposes the tools selected in this toolkit through the platform MCP server, so external MCP clients can call them.
6. Save Toolkit: Click Save to create the toolkit.

​

Available Tools:

​

Testing Toolkit Tools

1. Select LLM Model: Choose a Large Language Model from the model dropdown in the Test Settings panel
2. Configure Model Settings: Adjust model parameters like Creativity, Max Completion Tokens, and other settings as needed
3. Select a Tool: Choose the specific Teams tool you want to test from the available tools
4. Provide Input: Enter any required parameters or test queries for the selected tool
5. Run the Test: Execute the tool and wait for the response
6. Review the Response: Analyze the output to verify the tool is working correctly and returning expected results

​

Step 3: Add Teams Toolkit to Your Workflows

​

In Agents:

1. Navigate to Agents: Open the sidebar and select Agents.
2. Create or Edit Agent: Either create a new agent or select an existing agent to edit.
3. Add Teams Toolkit:
   • In the “TOOLS” section of the agent configuration, click the “+Toolkit” icon
   • Select your configured Teams toolkit from the dropdown list
   • The toolkit will be added to your agent with the previously configured tools enabled
   

​

In Pipelines:

1. Navigate to Pipelines: Open the sidebar and select Pipelines.
2. Create or Edit Pipeline: Either create a new pipeline or select an existing pipeline to edit.
3. Add Teams Toolkit:
   • In the “TOOLS” section of the pipeline configuration, click the “+Toolkit” icon
   • Select your configured Teams toolkit from the dropdown list
   • The toolkit will be added to your pipeline with the previously configured tools enabled
   

​

In Chat:

1. Navigate to Chat: Open the sidebar and select Chat.
2. Start New Conversation: Click + Conversation or open an existing conversation.
3. Add Teams Toolkit to Conversation:
   • Click the + icon in the chat input toolbar (tooltip: Add files, agents, toolkits and more…)
   • Hover over Toolkits in the popup menu
   • Search for your Teams toolkit by name
   • Enable the toolkit to add it to your conversation
4. Verify Toolkit Added: The Teams toolkit appears in the PARTICIPANTS panel on the right side under the Toolkits section.
5. Use Toolkit in Chat: You can now directly interact with your teams, channels and chats by asking questions or requesting actions that will trigger the Teams toolkit tools.

​

Instructions and Prompts for Using the Teams Toolkit

​

Instruction Creation for OpenAI Agents

• Direct and Action-Oriented: Employ strong action verbs and clear commands to initiate actions. For example, “Use the ‘list_chats’ tool…”, “Send a message to…”, “Post an update in the channel…”.
• Parameter-Centric: Clearly enumerate each parameter required by the tool. For each parameter, specify:
  • Its name (exactly as expected by the tool)
  • The format or type of value expected
  • How the Agent should obtain the value – whether from user input, derived from previous steps in the conversation, retrieved from an external source, or a predefined static value
• Contextually Rich: Provide sufficient context so the Agent understands the overarching objective and the specific scenario in which each Teams tool should be applied within the broader workflow. Explain the desired outcome or goal for each tool invocation.
• Step-by-Step Structure: Organize instructions into a numbered or bulleted list of steps for complex workflows. This helps the Agent follow a logical sequence of actions.
• Add Conversation Starters: Include example conversation starters that users can use to trigger this functionality. For example, “Conversation Starters: ‘Show me my recent Teams chats’, ‘Message the on-call engineer about the outage’, ‘Post the release notes in the Engineering channel’”

1. State the Goal: Begin by clearly stating the objective you want to achieve with this step. For example, “Goal: To notify the project team of a status update in their group chat.”
2. Specify the Tool: Clearly indicate the specific Teams tool to be used for this step. For example, “Tool: Use the ‘send_chat_message’ tool.”
3. Define Parameters: Provide a detailed list of all parameters required by the selected tool. For each parameter:
   • Parameter Name: <Parameter Name as defined in tool documentation>
   • Value or Source: <Specify the value or how to obtain the value. Examples: "user input", "from previous step", "hardcoded value 'General'", "value of variable X">
4. Describe Expected Outcome (Optional but Recommended): Briefly describe the expected result or outcome after the tool is successfully executed. For example, “Outcome: The Agent will confirm the message was sent and report the message_id and chat_id.”
5. Add Conversation Starters: Include example conversation starters that users can use to trigger this functionality.

You are a Teams collaboration assistant. Help the user keep track of chats and post updates on their behalf.

## Checking Recent Chats
- **Tool**: list_chats
- **Parameters**:
    - unread_only: true if the user only wants chats with unread messages
    - chat_type: "oneOnOne", "group" or "meeting" if the user specifies a type
    - limit: 50 (default, unless the user asks for more/fewer)
- **Process**:
    1. Call the tool and summarize each chat's topic/members, last message preview, and unread status
    2. Offer to open any chat in full with find_chat_messages

## Reading a Specific Chat
- **Tool**: find_chat_messages
- **Parameters**:
    - chat: chat ID, group chat topic, or the other person's email (for a 1:1 chat)
    - since / after_message_id: the watermark returned by the previous call, if any (omit on the first call)
    - senders: (optional) only messages from specific people
- **Process**:
    1. Report has_new / new_count and the messages, oldest first
    2. Store the returned watermark / latest_message_id for the next check

## Messaging People or a Chat
- **Tool**: send_chat_message
- **Parameters**:
    - chat or recipients: existing chat ID/topic, or the email(s) of the people to message
    - message: the text to send (confirm with the user before sending); write @[email] inside it to place an inline @mention where it belongs, e.g. "Thanks @[anna@contoso.com], please review"
    - mentions: (optional) people to @mention at the start of the message, for anyone not already placed inline
    - importance: (optional) 'high' or 'urgent' only when the user asks for it or the message is time-critical
- **Process**:
    1. Confirm the drafted message with the user before sending
    2. Send and report back the new message_id and chat_id (and whether a new chat was created)

## Posting to a Channel
- **Tool**: send_channel_message
- **Parameters**:
    - team: team ID or display name (from list_teams)
    - channel: channel ID or display name (from list_channels)
    - message: the text to post; same @[email] inline-mention syntax as send_chat_message
    - reply_to_message_id: (optional) set only when replying to an existing post
    - mentions: (optional) people to @mention at the start, for anyone not already placed inline
    - importance: (optional) 'high' only when the content is genuinely important
- **Process**:
    1. Confirm the drafted post with the user before sending
    2. Send and report back the new message_id and thread_id

## Conversation Starters
- "Do I have any unread Teams chats?"
- "Message Sarah to ask if the report is ready"
- "Post the sprint summary in the Engineering channel of the Platform team"
- "Search my Teams messages for anything about the outage"

​

Chat Usage Examples by Tool Category

User: "What teams am I a member of, and what channels does the Platform team have?"

Agent Response: [Agent uses list_teams tool, then list_channels tool with team="Platform"]

👥 **Your Teams (3)**:
1. **Platform**
2. **Marketing**
3. **Customer Success**

📁 **Channels in "Platform" (4)**:
1. **General**
2. **Engineering**
3. **Incidents**
4. **Releases**

User: "Show me my chats with unread messages."

Agent Response: [Agent uses list_chats tool with unread_only=true]

💬 **Unread Chats (2)**:

1. **Sarah Connor** (1:1) — last message: "Can you review the PR?" — 10 min ago
2. **"Release Planning"** (group, 5 members) — last message: "Pushed the new build" — 1 hour ago

User: "What did Sarah say in our chat since yesterday?"

Agent Response: [Agent uses find_chat_messages tool with chat="sarah@partner.com", since=<yesterday's timestamp>]

🧵 **Chat with Sarah Connor (2 new messages)**:

1. Sarah: "Can you review the PR when you get a chance?" (2026-10-01 15:02)
2. Sarah: "No rush, early next week is fine." (2026-10-01 15:03)

I've updated the watermark so the next check only shows messages after these.

User: "Search my Teams messages for anything about the outage."

Agent Response: [Agent uses search_teams_messages tool with query="outage"]

🔍 **Search Results for "outage" (3 matches)**:

1. Chat with On-Call Engineer — "Investigating the outage now" (2026-10-01)
2. Platform / Incidents channel — "Outage postmortem scheduled for Friday" (2026-09-30)
3. "Release Planning" group chat — "Rollback avoided the outage" (2026-09-29)

Would you like me to open the full thread for any of these? Channel hits can be read in full with search; chat hits can be opened with find_chat_messages.

User: "Message john@example.com and ask if the deployment is ready."

Agent Response: [Agent uses send_chat_message tool with recipients=["john@example.com"]]

✅ **Message Sent!**

- **To**: john@example.com
- **Message**: "Is the deployment ready?"
- **Message ID**: 1696262400000
- **Chat ID**: 19:abcdef...@unq.gbl.spaces
- **Chat Created**: false (existing 1:1 chat was used)

User: "Post the sprint summary in the Engineering channel of the Platform team, and mention Sarah."

Agent Response: [Agent uses send_channel_message tool with team="Platform", channel="Engineering", message="Sprint summary attached.", mentions=["sarah@partner.com"]]

✅ **Posted to Channel!**

- **Team**: Platform
- **Channel**: Engineering
- **Posted**: "@Sarah Connor Sprint summary attached."
- **Message ID**: 1696262400000
- **Thread ID**: 1696262400000

User: "Tell Anna in our chat that the build is green and mark it urgent."

Agent Response: [Agent uses send_chat_message tool with chat="anna@contoso.com", message="Thanks @[anna@contoso.com], the build is green.", importance="urgent"]

✅ **Message Sent!**

- **Posted**: "Thanks @Anna Kowalski, the build is green."
- **Importance**: urgent (Anna will be re-notified every 2 minutes for 20 minutes)
- **Message ID**: 1696262400001
- **Chat ID**: 19:abcdef...@unq.gbl.spaces

​

Troubleshooting

​

Support Contact

​

FAQ

https://login.microsoftonline.com/{tenant_id}