Salesforce Agent

Interact with Salesforce using a general-purpose AI agent that can create, update, and read objects like opportunities, contacts, and accounts based on a natural-language prompt.

This node has a preferred replacement: Structured Salesforce Agent. Use that node for new workflows, since it gives you explicit control over which objects and properties are touched instead of relying entirely on the agent's interpretation of a prompt.

Salesforce Agent node in the workflow builder

What It Does

Runs an AI agent against Salesforce using the prompt you provide. The agent can read, create, and update Salesforce objects on its own, and it responds in a conversation that AskElephant creates for the run.

When To Use It

Use this only for workflows already built on it, or for quick, low-stakes lookups where you do not need guardrails around exactly which properties get changed. For anything that writes data to Salesforce, especially deal or account fields, use Structured Salesforce Agent instead, since it lets you name the exact properties to update and require confirmation before syncing.

Configuration

Configure the node with these settings:

Field	What it controls
Related Object Type	The kind of object to use for context: `meeting`, `calendarEvent`, `person`, or `company`. Defaults to `meeting`.
Related Object	The specific record to use as context. All relevant information about it is provided to the agent.
Agent Prompt	The instructions for what the agent should do in Salesforce. Required.
Show Message	Whether to make your prompt visible in the resulting conversation. Defaults to off.
Show Completion	Whether to make the agent's response visible in the resulting conversation. Defaults to on.
Model	The language model the agent uses. Defaults to `claude-sonnet-4-5`.
Max Steps	The maximum number of tool call steps the agent can take before it must return a response. Defaults to 20.
Chat Title	An optional title for the created chat. If left blank, one is generated automatically.

Outputs

The node passes these values downstream. Check the Outputs tab on the configured node to confirm exact paths.

Variable	Type	Description	Example
`conversation`	`object`	A reference to the conversation where the message was sent	`{{nodeId.conversation}}`
`conversation.id`	`string`		`{{nodeId.conversation.id}}`
`conversationLink`	`string`	A clickable link to the created conversation in AskElephant	`{{nodeId.conversationLink}}`
`aiCompletion`	`string`	The LLM completion that was sent to the conversation	`{{nodeId.aiCompletion}}`

Example Workflow

After a call, a workflow prompts the agent to "find the opportunity linked to this account and summarize its current stage and next steps." The agent looks up the opportunity in Salesforce and posts its summary to a new conversation.

Structured Salesforce Agent
Update Opportunity
Search Opportunity by Stage

Need More Help?

Reach the AskElephant support team by:

clicking the Support tab on the right side of your screen,
emailing [email protected],
or using @askelephant support in your dedicated Slack channel.