--- schema_version: "1" id: a487933b67b899ec1daab726e42c78d300671e4d6a093e045bf8b8758630af21 title: Configuring Agent Operations for Agentforce Connector 1.3 collection: mulesoft version: current locale: en-us portal_url: https://docs.salesforce.com/mulesoft/current/en-us/agentforce-connector/latest/configuring-agent-operations.md canonical_url: https://docs.mulesoft.com/agentforce-connector/latest/configuring-agent-operations source_site: docs.mulesoft.com content_type: documentation content_hash: 132d1d489f2bb057d470bc67a4747d63d289d302c44607eecccb5412195aefb2 description: MuleSoft Documentation Site release: "1.3" latest: true --- --- product: Agentforce Connector version: 1.3 is-latest-version: true --- # Configuring Agent Operations for Agentforce Connector 1.3 > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Use these guidelines to configure the **Start agent conversation**, **Send Message (Sync)**, **Continue agent conversation (Deprecated)**, and **End agent conversation** operations for Agentforce Connector. ## Configure the Start Agent Conversation Operation The **Start agent conversation** operation establishes the connection between the app and the agent. This is the first step to enable an external application to interact with the AI agent. 1. Select the operation on the Anypoint Code Builder or Studio canvas. 2. In the **General** properties tab for the operation, enter these values: - **Display Name** Display name for the operation. - **Connector Configuration** Select the associated configuration for the connector. - **Agent list** Select the agent to invoke. The list shows only agents that are currently active in the org. The connector filters out any agents that aren’t in the `active` state. - **Bypass User** Select whether to use the agent-assigned user instead of the logged-in user. Setting this to `True` allows the agent to use its own permissions to access the necessary Data Cloud assets. This enables the agent to answer the user’s query even if you lack direct access. - **Agent Variables** Provide an array of custom or context variables to pass runtime values to the agent. Match each variable name to the corresponding variable defined in Agent Builder. ### Output The operation returns a session ID for the interaction between the app and the agent. The session ID is used as an input for the **Continue agent conversation** operation. ## Configure the Send Message (Sync) Operation The **Send Message (Sync)** operation sends a message from the MuleSoft application to the agent and waits for a synchronous response. The agent maintains the context of the conversation and provides a relevant response using a structured JSON payload that separates business data from metadata. This operation replaces the deprecated [**Continue agent conversation**](#continue-agent) operation to provide full access to all elements of the response payload. 1. Select the operation on the Anypoint Code Builder or Studio canvas. 2. In the **General** properties tab for the operation, enter these values: - **Display Name** Display name for the operation. - **Connector Configuration** Select the associated configuration for the connector. - **Message** Enter plain text instructions to help the agent make decisions for different use cases. - **Session ID** Enter the session ID that was returned with the **Start agent conversation** operation. This is required to send the prompt defined in the **Message** field to the agent. - **Message Sequence Number** Enter `1` for the first message in a session, and increment the value by `1` for each subsequent message to preserve message order. - **Agent Variables** Provide an array of custom or context variables to pass runtime values to the agent. Match each variable name to the corresponding variable defined in Agent Builder. ## Configure the Continue Agent Conversation Operation (Deprecated) The **Continue agent conversation** operation surfaces the message from the MuleSoft application to the agent. The agent maintains the context of the conversation and provides a relevant response (or actions) according to the user’s input. The **Continue Agent Conversation** operation is now deprecated. Use the [**Send Message (Sync)**](#SendMessageSync) operation instead. 1. Select the operation on the Anypoint Code Builder or Studio canvas. 2. In the **General** properties tab for the operation, enter these values: - **Display Name** Display name for the operation. - **Connector Configuration** Select the associated configuration for the connector. - **Message** Enter plain text instructions to help the agent make decisions for different use cases. - **Session ID** Enter the session ID that was returned with the **Start agent conversation** operation. This is required to send the prompt defined in the **Message** field to the agent. ## Configure the End Agent Conversation Operation Use the **End agent conversation** operation to close the session after the agent responds. > [!NOTE] Sessions are closed automatically after three days of inactivity. 1. Select the operation on the Anypoint Code Builder or Studio canvas. 2. In the **General** properties tab for the operation, enter these values: - **Display Name** Display name for the operation. - **Connector Configuration** Select the associated configuration for the connector. - **Session ID** Enter the session ID that was returned with the **Start agent conversation** operation. ## See Also - [Using Anypoint Code Builder to Configure a Connector](https://docs.mulesoft.com/connectors/introduction/intro-config-use-acb) - [Using Studio to Configure a Connector](https://docs.mulesoft.com/connectors/introduction/intro-config-use-studio) - [Agentforce Connector Reference](https://docs.mulesoft.com/agentforce-connector/latest/agentforce-connector-reference) - [Salesforce Help](https://help.salesforce.com)