Welcome back to the FME Knowledge Roundup! The Q2 edition brings a fresh batch of knowledge articles and tutorials spanning MCP integrations, cursor-based API pagination with FME's new native looping, and Snowflake — from Variants to running Remote Engines directly inside Snowpark.
We've also got updates on FME Flow fundamentals and webhook networking, security hardening in 2026.2, working with serverless cloud infrastructure, raster alpha compositing, and automating work orders in Trimble Unity. As always, we've rounded up the most essential reads this quarter to help you get the most out of FME.
In This Issue:
Here's what you'll find in this quarter's Must-Read Picks:
MCP
Six new articles cover Model Context Protocol with FME.
On the consuming side, the MCPCaller transformer lets a workspace connect to external MCP servers, list their tools, and call them with structured JSON input.
On the serving side, FME Flow can publish existing workspaces as MCP tools that AI agents and other MCP clients can discover and run, with dedicated guides on returning text and image content back to the calling client via the MCP Writer.
- Getting Started with MCP in FME
- Getting Started with FME Flow as an MCP Server
- Getting Started with the MCPCaller
- Getting Started with the MCPCaller: List Tools and Call Tools
- Returning Image Content from an FME MCP Tool
- Returning Text Content from an FME MCP Tool
AI in FME
Two new articles show AI vision models in action: one automates QA for field asset inspections by pairing ArcGIS Survey123 with AI-derived condition ratings and weekly PDF reports. The other article covers extracting structured data from images, like meter readings and drawing metadata, using the GoogleGeminiConnector.
- Automated QA for Field Asset Inspections Using AI, Webhooks, and FME
- Getting Started with AI in FME: Image and Visual AI Tools
Data Virtualization
Two new tutorials expand on building Data Virtualization APIs: one covers creating a PUT endpoint to update existing records (with path parameters, schema validation, and 200/400/404 responses), while the other walks through handling POST requests and file uploads for formats without explicit MIME mapping, like .fmw workspace files.
- Creating a PUT Endpoint in a Data Virtualization API
- Handling Data Virtualizing POST Requests and File Uploads for Formats Without Explicit MIME Mapping
FME Flow
Three articles round out your FME Flow toolkit: a refreshed overview of FME Flow's architecture, components, and deployment options for anyone getting started; a deep dive into the networking and firewall considerations behind Survey123-to-FME Flow webhook integrations; and a practical guide to choosing the right email mechanism in FME Flow.
- Introduction to FME Flow
- Networking Considerations for FME Flow Webhook Integrations
- Best Practices for Using Email in FME Flow
Looping
An updated tutorial walks through cursor-based API pagination using FME 2026.2's new native looping functionality, building a loop with the HTTPCaller, JSONFragmenter, TestFilter, and JSONFlattener to recursively pull all messages from Slack's conversation history API.
Rasters
A new tutorial shows how to fade an orthophoto to white at its project boundary instead of clipping it abruptly, using buffer rings, an ImageRasterizer-built alpha mask, and RasterMosaicker's alpha-band compositing.
Remote Engines
A new tutorial tackles a known Remote Engines Service limitation — job log details live only on the remote engine host, invisible from the FME Flow core UI — by building a Workspace App that checks a user's job permissions via the Data Virtualization API, then retrieves and returns the corresponding remote engine log directly from the core.
Security
FME 2026.2 introduces security tiers, giving admins control over which powerful capabilities (like advanced scripting) are permitted in their environment. This article breaks down the new Default Hardening and Minimal Hardening tiers, how to change them in FME Form or FME Flow, and what to do if a workspace hits a restricted capability.
Serverless Data
A new overview article maps FME's capabilities to serverless cloud infrastructure — connecting to object storage (S3, Azure Blob, GCS), managed databases (DynamoDB, Snowflake, BigQuery), event-driven Automations, and cloud-native formats like GeoParquet and COG — with real customer examples of FME running pipelines entirely within AWS, Azure, and Google Cloud.
Snowflake
Five resources now cover FME and Snowflake end to end: a hub article for getting connected and working with Snowflake data; a guide to reading, writing, and querying JSON stored in Snowflake Variant columns (including a full BIM/IFC example); and three articles on running FME directly inside Snowflake via the Remote Engine Service.
- Getting Started with Snowflake
- Using FME Remote Engine Service Inside Snowflake
- Working with JSON Snowflake Variants in FME
- Authoring for Snowflake Remote Engines
- Troubleshooting Remote Engines Service in Snowpark
TrimbleUnityConnector
A new tutorial walks through automating Work Order creation in Trimble Unity (or Cityworks) from an Excel source using the TrimbleUnityConnector, then extends the workflow with the HTTPCaller to hit API endpoints the connector doesn't expose directly, like posting comments to a newly created Work Order.
2026.2 Help Documentation Highlights
FME Form
- Improved help topics: We've updated and refined 40+ connector-related transformer help topics. Open the Transformer List and filter by Connector.
- Looping Transformers: Looping Transformers support iterative workflows.
- Refined handling of unavailable transformers: Updated behavior for missing or unavailable transformers.
- Model Context Protocol support: We added an MCP Writer and an MCPCaller transformer.
- Updated AI Assist: Enhancements to AI Assist.
FME Flow
- MCP Server support: Learn about working with MCP Servers.
- SFTP Directory trigger: The SFTP Directory trigger monitors activity on an Secure File Transfer Protocol (SFTP) server, and sends JSON messages about that activity.
- Updated Flow Apps security model: A revised security model for Flow Apps.
Let us know your thoughts!
We hope this edition of the FME Knowledge Roundup helps you stay ahead of the latest platform updates and best practices!
Have suggestions for next quarter? Let us know what topics you'd like us to cover, we build these roundups with your feedback in mind.