Quick Start: End-to-End Lifecycle Workflow
Use this page as the runbook for executing FabricOps Starter Kit in Microsoft Fabric.
Prerequisites
- Python 3.11+ and
uvinstalled locally. - Access to a Microsoft Fabric workspace with Lakehouse/Warehouse and Environment permissions.
- A dataset with defined business purpose, ownership, and acceptance rules.
1) Install and prepare
git clone https://github.com/Voycepeh/FabricOps-Starter-Kit.git
cd FabricOps-Starter-Kit
uv sync --all-extras
uv build
For Fabric Environment install steps, use Fabric Wheel Install.
2) Open your Fabric notebook and load/create config
- Start from
templates/notebooks/fabric_data_product_mvp.md. - Attach the Fabric Environment that contains the package wheel.
- Create or load runtime/framework config (paths, dataset IDs, output targets, run context).
3) Execute in canonical 10-step order
Follow this sequence (details in Lifecycle Operating Model):
- Define purpose, approved usage, and governance ownership.
- Configure runtime, environment, and path rules.
- Declare source contract and ingest source data.
- Validate source against contract and capture metadata.
- Explore data and capture transformation / DQ rationale.
- Build production transformation and write target output.
- Validate output and persist target metadata.
- Generate, review, and configure DQ rules.
- Generate and review classification / sensitivity suggestions.
- Generate data lineage and handover documentation.
4) Validate completion criteria
A successful run should leave you with:
- source and output profiling artifacts,
- quality and drift results,
- written output table(s),
- governance metadata and reviewer decisions,
- lineage output,
- handover documentation artifacts.
5) Next references
- Canonical lifecycle detail: Lifecycle Operating Model
- Notebook conventions: Notebook Structure
- Architecture: Architecture
- Callable catalogue: Reference