SnapLogic MCP Server tools FAQ

Common questions about client compatibility, licensing, PyGen text-to-pipeline generation, and SLPy.

Which AI clients can connect to the SnapLogic MCP Server tools?

Any MCP-compatible client can connect, including Claude Code, Claude Desktop, VS Code Agent Mode, GitHub Copilot, and custom agents. The MCP Server implements the standard MCP protocol over Streamable HTTP; SnapLogic requires no client-specific integration. For setup examples, refer to Connect an MCP client.

Do the MCP Server tools require a special license or feature flag?

The Platform and PyGen toolsets are available with no additional license or feature flag.

Note: The MCP toggle in the Pipeline Properties dialog is a separate feature that lets you expose your own pipeline as an MCP-compatible endpoint. It has no effect on the standalone MCP Server tools documented here.

Which toolsets are available, and what does each provide?

All three paths are supported — /mcp (full toolset), /mcp/platform (platform operations), and /mcp/pygen (PyGen text-to-pipeline tools). Refer to SnapLogic MCP Server tools for a description of each path.

Can SnapGPT connect to the MCP Server as an MCP client?

SnapGPT does not support connecting to the SnapLogic MCP Server as an MCP client. For AI-assisted pipeline creation within SnapLogic, use SnapGPT directly in the Designer.

PyGen is described as read-only — how does text-to-pipeline generation work in practice?

PyGen handles the knowledge side: it looks up Snap definitions, retrieves pipeline examples in SLPy (SnapLogic Python) format, and validates pipeline structure. It doesn't create pipelines directly. Your MCP client assembles the SLPy pipeline definition from those responses and then calls import_pipeline (Platform toolset) to create the actual pipeline in SnapLogic. Refer to PyGen toolset for the full list of available tools.

What clients support text-to-pipeline generation, and what does "SLPy-aware" mean?

SnapCode (Claude Code plugin) and SnapGPT (Designer) are the two supported entry points for text-to-pipeline generation. Both are SLPy-aware — meaning they understand SLPy structure (pipeline declarations, Snap configurations, and Snap connections) well enough to produce valid output from PyGen responses. SLPy is an internal format with no public schema or standalone SDK; there's no supported path for making a third-party agent SLPy-aware today.

Is PyGen enabled at the account level or the environment level?

PyGen is configured per environment, not at the account level. If /mcp/pygen returns 404, or /mcp shows only 34 tools with no PyGen tools, PyGen isn't enabled for that environment. Contact your SnapLogic administrator to enable it.