Skip to content

Product Roadmap

This document outlines the strategic development milestones and feature releases planned for fme-mcp.

Version 1.0.0 (Released — November 2026) ✅

  • Core MCP Server: Streamable HTTP and stdio transports based on MCP specification 1.x.
  • Universal FME Engine (2018.0 - 2026+):
  • Direct local FME Form execution via fme.exe CLI and fmeobjects bindings.
  • FME Flow REST API client with full authentication and workspace discovery.
  • Security Engine:
  • Path traversal guard with canonical directory allowlists.
  • Command injection guard rejecting shell metacharacters.
  • SSRF protection against private IPv4/IPv6 ranges and cloud metadata services.
  • Sensitive parameter masking and redaction.
  • Role-Based Access Control (RBAC).
  • Core Tool Suites:
  • Workspaces & Jobs execution and inspection.
  • Spatial analysis, coordinate reprojection, and geometry buffer operations.
  • Data engineering, profiling, quality validation, and schema inference.
  • Base Adresse Nationale (BAN) address geocoding and automated enrichment.
  • Distribution:
  • Open-source release under Apache 2.0.
  • Automated CI/CD GitHub Actions pipeline.
  • Published on GitHub Container Registry and GitHub Releases.

Version 1.1.0 (Target: Q1 2027) 🔄

  • FME 2026 Parity: Native emulation of Safe Software 2026 MCP server protocols and JSON-RPC endpoints.
  • Asynchronous Webhook Reception: Inbuilt HTTP webhook receiver for real-time FME Flow notification callbacks.
  • Enhanced Vector Indexing: Automatic semantic indexing of workspace documentation and custom transformers using local embeddings.
  • Extended Spatial Tools: Advanced 3D city model and BIM/IFC inspection tools via FME Quick Translator bindings.

Version 1.2.0 (Target: Q2 2027) 📋

  • Enterprise Governance Suite: Automated lineage graphing across multi-workspace data pipelines.
  • Federated MCP Gateway: Downstream routing and composition across external MCP servers (PostGIS, GeoServer, QGIS, ArcGIS Enterprise).
  • High-Throughput Batch Processing: Celery and Redis distributed job queue integration for multi-worker FME clusters.