Connector nodes

What they are

Connector nodes let you interact with third-party services directly from the canvas, with no manual HTTP configuration. Each connector provides pre-built nodes for a specific service (e.g., GitHub, Salesforce, Slack, Microsoft 365), with typed inputs, typed outputs, and automatic authentication through Integration Service.

Connector nodes appear in the node palette under the Connector category, grouped by service.

How they work

Connector nodes are loaded dynamically from UiPath's Integration Service. The available connectors depend on what's configured in your Automation Cloud organization. Each connector node maps to a specific API operation on the target service.

When you add a connector node to the canvas, Flow fetches its input schema and renders a dynamic configuration form in the properties panel. You fill in the required fields, select a connection for authentication, and the node handles the API call at runtime.

Configuration

Connexion (Connection)

Every connector node requires a connection, a pre-configured authentication credential managed in Integration Service. The connection is selected in the properties panel when configuring the node.

Connections handle token refresh, OAuth flows, and credential storage automatically. Auth headers and secrets are not managed directly.

Remarque :

Connections are configured in the UiPath Automation Cloud portal under Integration Service. Refer to the Integration Service documentation for setup instructions.

Node-specific inputs

Each connector node has its own input fields based on the API operation it wraps. The properties panel renders these dynamically. You'll see different fields for a "Create Issue" node than for a "List Repositories" node, even within the same connector.

All input fields support variable expressions ($vars.<nodeName>.<property>).

Sortie

Connector node output is available at $vars.<nodeName>.output. The shape of the output depends on the specific node and the API it calls.

// Access the output of a GitHub connector node named "listBranches1"
$vars.listBranches1.output

Gestion des erreurs

Connector nodes support error handles. If the API call fails and an error handle is connected, execution routes to the error path with the standard error object at $vars.<nodeName>.error.

Refer to Error handling for details.

Connector triggers

Some connectors also provide triggers: start events that fire when something happens in the external service (e.g., a new email arrives, a form is submitted). Connector triggers appear in the Triggers section of the node palette.

Trigger inputs are owned by the trigger and accessed as $vars.<triggerName>.output.<inputName>. Refer to Triggers for details.

When to use a connector vs HTTP Request

Use a connector node when...	Use HTTP Request when...
A connector exists for your target service	No connector exists for the service
You want typed inputs and outputs with no manual configuration	You need full control over headers, query parameters, and body
You want automatic authentication via Integration Service	The API uses a non-standard auth scheme
You want a maintainable process that adapts to API changes	You're prototyping against a new or internal API

Rule of thumb: The node palette is the first place to check for a connector. HTTP Request is the fallback when no connector exists for the target service.

Available connectors

The full list of available connectors depends on your Automation Cloud organization and Integration Service configuration. For connector-specific documentation, including setup, authentication, and available nodes, refer to the UiPath Integration Service documentation:

Integration Service connectors documentation

Pages liées

• HTTP Request node: general-purpose API calls when no connector exists
• Triggers: connector-based triggers and other trigger types
• Error handling: handling failures from connector nodes
• Variables and data flow: accessing connector output with $vars