Use SnapCode

Start a SnapCode session

Open a terminal in any directory and run:

claude

Claude Code opens and SnapCode is ready. You interact entirely through the chat, no commands or syntax to memorize.

Build a pipeline

Enter a natural language prompt in English to describe your pipeline. Be specific about the data source, the transformation, and the destination to get the best results without follow-up questions.

Example:

Build a pipeline that reads customer orders from Salesforce,
filters cancelled ones, and writes the results to Snowflake

SnapCode generates the pipeline source file (.py) and translates it to the Platform format (.slp). For transformations that can run locally, it also executes the pipeline against local test data before deployment.

If SnapCode needs clarification, such as the Salesforce object to read, it asks a question. Answer it and SnapCode continues automatically.

Deploy a pipeline

Inform SnapCode where to deploy the pipeline:

Deploy it to the SLDemo org, projects/your-name/your-project path

The first time you deploy in a session, SnapCode asks for:

• Environment/Org: Your SnapLogic organization name (for example, snaplogic or SLDemo)
• Project path: Path to store the pipeline (for example, projects/your-name/your-project)
• Snaplex: Runtime to use for validation

Before deploying, SnapCode shows the plan (org, project path, and Snaplex) and waits for you to confirm. SnapCode prompts to save these for the rest of the session. You can also include all three upfront in your deploy message to skip the prompts.

SnapCode uploads the pipeline and validates it on the SnapLogic Platform. If validation fails, it displays the error and automatically fixes it before redeploying, up to 5 times.

Deploy child_pipeline.slp, then parent_pipeline.slp to...

Run a pipeline

Run it

SnapCode triggers the pipeline on the SnapLogic Platform and returns the execution status, including per-Snap error details if it fails.

Edit a pipeline

You can change a deployed pipeline in two ways:

• Conversationally: Ask SnapCode to make the change, for example:

  Fix the Mapper expression — subtotal should be qty × unit_price

  SnapCode edits the .py file and retranslates it. When you ask it to deploy, it shows the plan for you to confirm, then validates.
• By editing the file directly: Open the generated .py file in any text editor, make your changes, save, then prompt SnapCode:

  Deploy it

  SnapCode picks up your edited file and redeploys from it.

Validate an existing pipeline

To validate a pipeline on the SnapLogic Platform without running it:

Validate pipeline.slp

SnapCode validates the file on the Platform and reports any errors with the Snap name and field that caused the issue.

Use a local file as pipeline input

To upload a local file to SLDB so a pipeline can read it:

Upload sales_data.csv to SLDB at /snaplogic/my-project/

Then reference that path in your pipeline description when building.

Build a reusable expression library

Expression libraries let you define helper functions once and reuse them across pipelines.

Create a date helpers expression library —
today's date, previous month start/end, ISO formatting

SnapCode generates a .expr file, and uploads it to SLDB when you deploy a pipeline that uses it.

Where your files are stored

All generated files (.py, .slp, .expr) are created in your current working directory on your local machine. They persist across sessions. They are regular files on disk.

To keep the generated files organized, you can:

• Prompt SnapCode at the start of a session to put files in a subfolder:

  Put all generated files in my-pipelines/salesforce-project/

• Install SnapCode and run claude from any project directory of your choice. The plugin makes SnapCode available from any directory. Refer to Install SnapCode.

Known issues