Skip to main content
Version: 2.3

MaestroHub 2.3.0 Release Notes

MaestroHub 2.3.0 adds four new connectors (Azure IoT Hub, Google Pub/Sub, Databricks SQL, Databricks Storage), a centralized audit trail for compliance tracking, a redesigned overview page with health monitoring, a streamlined MCP toolset consolidated from 81 to 21 tools, pipeline diagnostics for structural issue detection, and broad improvements across dashboards, security, and UI polish.

Highlights​

  • 4 new connectors — Azure IoT Hub, Google Pub/Sub, Databricks SQL, and Databricks Storage.
  • Centralized audit trail — Compliance and security monitoring across all modules with per-resource actor attribution.
  • MCP tool consolidation — 81 tools reduced to 21 with OAuth2 scope enforcement and enriched ECharts metadata.
  • Redesigned overview page — Health command center with runtime metrics, welcome flow, and skip option for returning users.
  • Pipeline diagnostics — Detect structural issues in pipelines before execution.
  • UNS dashboards phase 3 — Variable support in panel titles, dashboards refactor phase four.

New Connectors​

Azure IoT Hub​

  • Send device-to-cloud and cloud-to-device messages over AMQP.
  • Supports connection string and shared access policy authentication.
  • Dedicated Azure IoT Hub pipeline nodes for read and write operations.

Google Pub/Sub​

  • Publish and consume messages from Google Cloud Pub/Sub topics and subscriptions.
  • Supports service account JSON key authentication.
  • Dedicated Pub/Sub pipeline nodes under the Cloud category.

Databricks SQL​

  • Query and execute statements on Databricks SQL warehouses.
  • Supports personal access token and OAuth2 M2M authentication.
  • Function types include SQL Query and SQL Execute with full parameter binding.
  • Dedicated Databricks SQL pipeline nodes under the Database category.

Databricks Storage​

  • Read and write files on Databricks Unity Catalog volumes.
  • Supports personal access token and OAuth2 M2M authentication.
  • Function types include Volume Read, Volume Write, Volume List, and Volume Delete.
  • Dedicated Databricks Storage pipeline nodes under the Storage category.

Centralized Audit Trail​

  • Audit trail module — Tracks create, update, delete, and execute actions across all modules with timestamps, actor identity, and resource details.
  • Per-resource attribution — All 26 audit event handlers consistently populate resource names and resolve actor emails via user lookup.
  • Auth integration — Logout events are no longer skipped, ensuring complete audit trail coverage.

MCP Engine Improvements​

  • Tool consolidation — Connectors, engine, and UNS tools reduced from 81 to 21, simplifying the MCP interface for AI assistants.
  • OAuth2 scope enforcement — All MCP tool handlers validate OAuth2 scopes, preventing unauthorized access.
  • ECharts metadata — Custom ECharts panel definitions enriched with data shape documentation and examples for better AI-assisted dashboard creation.
  • Connection status history — Connection status history now available as an MCP tool.

Pipeline Engine Improvements​

  • Pipeline diagnostics — New structural issue detection identifies configuration problems before pipeline execution.
  • Trigger node guidance — New pipeline users are guided to start with a trigger node.
  • Manual trigger fix — Resolved double-wrapping of manual trigger payloads.
  • Execution replay — Tabs now sorted by pipeline topological order for consistent navigation.
  • Change detection — Node Description and Labels are now included in substantive change detection.

UNS Dashboard Enhancements​

  • Dashboards refactor phase four — Continued improvements to dashboard architecture and rendering.
  • Panel title variables — Support for variables in panel titles for dynamic, context-aware dashboards.
  • Background images — Images are now stored as-is instead of being resized to 1920x1080.
  • Lite retention — Increased data retention to 10 minutes for Lite edition.
  • SQLite migration — Renamed tags column to labels for consistency.

UI/UX Improvements​

  • Redesigned overview page — New health command center with runtime/system metrics (CPU, RAM, disk, goroutines) and a welcome flow for first-time users.
  • Welcome page skip — Returning users can skip the welcome page with a new skip option.
  • Admin creation flow — Informative strip added to guide users during initial account setup.
  • Deleted entities — View deleted entities with role-based access control.
  • Data explorer — Clear messages button added to the live messages section.
  • Organization ID — Now visible and copyable across organization pages.
  • Execution history — Copy-to-clipboard support for execution details.
  • Open source libraries — Library list displayed in the UI with auto-updated notices on build.
  • Toolbar cleanup — Action buttons hidden when list is empty.
  • Sidebar spacing — Increased menu item spacing for better readability.
  • Dark theme — Login page now respects dark theme setting.
  • Dialog fix — Pointer-events cleared on body when dialog closes.

Auth & Security​

  • PAT org membership validation — Personal access tokens now validated against organization membership with unified HTTP/MCP middleware.
  • JWT org claims — Organization memberships embedded in JWT tokens for stateless validation.
  • OAuth2 client credentials — Org claim now embedded in client_credentials tokens for external OAuth2 clients.
  • PAT scope flexibility — PAT creation allowed with empty scopes when RBAC is disabled.
  • Org header enforcement — X-Organization-ID header now required on all requests, removing previous exemptions.
  • OPC UA certificate auth — Certificate authentication with auto-generated certificate support.
  • Dependency updates — Vulnerable Go dependencies updated (Trivy scan), SAST findings addressed in both backend (gosec) and frontend.

Connectors​

  • ConfigPreparer interface — New pre-save config mutation hook for connectors.
  • Modbus reconnection — Fixed automatic reconnection being blocked after connection loss.
  • Org-id validation — Added to connectors runtime and execution services.
  • Column widening — Connectors org_id columns widened from VARCHAR(36) to VARCHAR(255).
  • Connection/function name resolution — Frontend now always resolves names from API instead of storing in pipeline JSON.

Operations & Reliability​

  • License reset — Trial license now resets on every restart instead of persisting state.
  • UNS config — Packaging updated to match post-refactor structure.
  • Search index — Deleted pipelines and models now correctly removed from search index.
  • Timeline sort — Fixed sub-millisecond timestamp sorting in execution timeline.
  • License badge — Edition badge displayed in uppercase on the license page.

Getting Started​

Download a native binary or pull the Docker image and follow the Getting Started guide to have MaestroHub running in minutes.