Introduction
The MCPCaller lets an FME workspace call tools hosted by external MCP servers.
This means a workspace can connect to systems that expose an MCP interface without needing a dedicated FME connector for each system. The workspace calls the external tool, provides the required inputs, and receives the result as part of the workspace run.
As more platforms adopt MCP, those systems become reachable from FME through the same transformer. A workflow can use MCPCaller to pull data from another system, trigger an action, run a validation, or invoke processing that happens outside FME.
The returned result can then be used directly in the workspace logic, just like output from any other transformer or reader.
Why Use the MCPCaller?
Call external tools without building a custom connector
Any system that exposes an MCP interface becomes accessible from within an FME workflow through the MCPCaller, without writing a custom integration for each one.
Enrich workflows with data from connected systems
The MCPCaller can fetch data from an external MCP server mid-workflow, combining the results with data already in the workspace before writing to a destination.
Trigger actions in external systems during a translation
The MCPCaller can invoke write operations on external systems as features are processed, updating records or triggering processes without a separate integration.
Select tools dynamically based on incoming data
A workspace can discover available tools on an MCP server at run time and invoke the appropriate one based on the data being processed, without hardcoding a specific tool call.
Chain multiple MCP tool calls in a single workflow
Multiple MCPCallers can be used in sequence, passing outputs from one tool call as inputs to the next, orchestrating operations across different MCP-enabled systems within a single workspace.
Access MCP-exposed resources and prompts
Beyond tools, the MCPCaller can read resources and use prompts from external MCP servers, making files, records, documents, and reusable templates available within a workflow.
Key Concepts
Two Ways to Use the MCPCaller
The MCPCaller can be used in two ways, depending on whether the tool to call is known at design time or determined at run time.
Calling a specific tool directly
The MCPCaller is configured to call a designated tool on the MCP server. Workflow input data is passed as inputs, and the result is returned as structured output.
Selecting a tool based on input
An AI component evaluates the incoming data and selects the appropriate tool before passing the call to the MCPCaller. This allows a single workspace to route to different tools without hardcoding the choice.
Connecting to an MCP Server
The MCPCaller requires the URL of an MCP server and any required authentication credentials, configured in the transformer parameters. Tools, Resources, and Prompts are all supported.
- The tool, resource, or prompt to invoke can be set as a fixed value or driven by a feature attribute.
- When a tool call fails, the MCPCaller routes features to a fallback output port. This can be configured to handle errors without failing the workspace.
For full configuration details, see the MCPCaller documentation.
Articles
Getting Started with the MCPCaller: List Tools and Call Tools
Step-by-step guide to adding the MCPCaller to a workspace, connecting to an MCP server, and configuring parameters in design-time and run-time modes.
Additional Resources
Getting Started with FME in MCP: An introduction to MCP in FME — covers what MCP is, how FME supports it as both a client and a server, and links to the MCPCaller and FME Flow MCP Server articles.
FME with MCP: The Power of Choice, Expanded: Overview of FME's MCP capabilities, including the MCPCaller Transformer and FME Flow's MCP Server, with use cases and FAQs.
The Model Context Protocol: A Universal Bridge Between Your Data and AI: Conceptual overview of MCP — what it is, how it works, and how FME fits into the ecosystem as both a consumer and provider of MCP capabilities.
MCP and the Power of Choice: Expanding Your AI and System Reach (webinar): Introductory walkthrough of both FME MCP capabilities with live demos — consuming MCP tools with the MCPCaller and creating MCP tools in FME Flow.
From Esri to AI: Create an MCP Server in 7 Minutes with FME (webinar): Live build turning an FME workspace into a functional MCP server, using ArcGIS data as a starting point. Covers the full pattern from workspace to MCP tool.
From Prototype to Production: Building Real-World AI and MCP Workflows (webinar): Intermediate-level session with real municipal use cases, covering generative AI for document extraction, the FME Flow MCP server in practice, and lessons learned from production deployments.