What problem does it solve? Promoting Microsoft Fabric content across dev, test, and production stages requires many manual portal clicks, and the Fabric REST surface has non-obvious pitfalls: deploys are long-running operations, IDs must be resolved by name, permissions span two systems, and there is no compare API to detect changed items. This Skill drives the entire deployment-pipelines lifecycle from an agentic CLI using az rest and curl. ## Core Features & Use Cases - Pipeline and Stage Lifecycle: Create, update, and delete deployment pipelines with 2-10 ordered stages, and assign or unassign workspaces to stages with idempotent scripting. - Stage-to-Stage Deployment: Deploy all supported items or a selective items list forward or backward, then poll the long-running operation to a terminal state and report per-item results. - Changed-Item Detection: Compute the delta between stages yourself using the bundled diff_item_definitions.py script, which normalizes auto-rebound binding fields so unedited items do not appear changed. - Use Case: A user asks to promote a validated Sales dev workspace to the Test stage. The Skill resolves pipeline, stage, and workspace IDs by name via JMESPath, POSTs the deploy with a note, polls /v1/operations/{id} until completion, and reports which items moved. ## Quick Start Ask the assistant to create a three-stage Fabric deployment pipeline named Sales Analytics ALM and assign your development workspace to its first stage.
