Create a pipeline

In SnapLogic, you can create the following types of pipelines:

Standard pipeline: The default pipeline type that runs once per invocation to process and integrate data from start to finish.
Resumable: Suspends instead of failing when an endpoint goes down mid-execution, allowing you to resume processing from the exact point of interruption.
Cache: Stores up to 100,000 reference data records in memory for fast, repeated lookups within a pipeline —eliminating costly round-trips to external systems.
MCP: Exposes SnapLogic pipelines as discoverable AI tools via the Model Context Protocol, enabling any MCP-compatible AI agent or client to invoke them seamlessly.
Tools: Pipelines exposed as callable tools that can be invoked directly by an Agent pipeline or through an MCP Server.

To create a new pipeline:

In the Designer canvas, click the Add icon or click the Pipeline icon to open the Add New Pipeline dialog.
Specify a name for your pipeline in the Label field. The names of assets, projects, or project spaces are limited to UTF-8 alphanumeric characters and special characters.
Select the Project path to save your pipeline. If you plan to save this pipeline to a different project from the one you last accessed, change the path of the target project.
If this pipeline is a tool to be called by an Agent or MCP Server, enable the Tool access toggle in the Tools tab.

Note: Agent and MCP Server pipeline types are set automatically when the corresponding Snaps are added to the canvas and cannot be manually set.
To make this a resumable pipeline, select Resumable mode checkbox.
Configure the remaining settings under the following tabs:
- Parameters
- Tools
- OpenAPI
- Info
Click Save. You now have a blank canvas to build your pipeline.

You can use the Snap Reference documentation to do the following:

To configure pipeline settings such as parameters, tool inputs, and expression libraries, see Pipeline properties and configuration.