adf-master

Generate ADF pipeline JSON with proper structure, activities, and best practices

Debug ADF pipeline failures, analyze error messages, and suggest fixes

Generate ADF expressions for date manipulation, string operations, and dynamic content

Generate ADF linked service JSON for any connector with proper authentication

Validate ADF pipeline JSON files for nesting violations, resource limits, and configuration issues

Use this agent when the user needs help with Azure Data Factory pipeline JSON, activities, linked services, datasets, triggers, expressions, validation, CI/CD, Microsoft Fabric integration, Databricks orchestration, or ML pipeline patterns. <example> Context: User needs to create an ADF pipeline user: "Create a pipeline JSON that copies data from Azure SQL to Blob Storage as Parquet" assistant: "I'll create a complete pipeline JSON with the Copy activity, linked services, and datasets configured for Azure SQL source and Blob Storage Parquet sink." <commentary>Triggers for any ADF pipeline JSON creation or editing</commentary> </example> <example> Context: User has a validation issue user: "Can I put a ForEach inside another ForEach in ADF?" assistant: "No, ADF prohibits nested ForEach. The workaround is to use Execute Pipeline activity inside the outer ForEach to invoke a child pipeline containing the inner ForEach." <commentary>Triggers for ADF validation rules and nesting limitations</commentary> </example> <example> Context: User needs Fabric integration user: "How do I copy data to Microsoft Fabric Warehouse from ADF?" assistant: "I'll provide the linked service and dataset JSON for Fabric Warehouse using the dedicated connector with managed identity authentication." <commentary>Triggers for Microsoft Fabric / OneLake integration</commentary> </example> <example> Context: User needs ML orchestration user: "How do I call an Azure ML batch endpoint from an ADF pipeline?" assistant: "I'll show you the WebActivity pattern for invoking Azure ML batch endpoints with managed identity, including the request body format and output handling." <commentary>Triggers for ML pipeline orchestration and batch scoring</commentary> </example> <example> Context: User needs expression help user: "How do I format a date in an ADF expression for a file path?" assistant: "Use @formatDateTime(utcnow(),'yyyy/MM/dd') to create a date-partitioned path. I'll show the full expression with concat for the complete path." <commentary>Triggers for ADF expression language functions</commentary> </example>

This skill should be used when the user asks about "Azure Data Factory CI/CD", "ADF ARM template deployment", "ADF npm build validation", "PrePostDeploymentScript", or needs guidance on ADF pipelines, triggers, GitHub Actions, or Azure DevOps integration.

This skill should be used when the user asks about "Azure ML batch endpoints in ADF", "Azure OpenAI Batch API pipeline", "ADF ML scoring", "SQL to Storage archival", or needs guidance on AI Services integration, Databricks ML training, or Data Flow feature engineering.

This skill should be used when the user asks about "ADF activity nesting rules", "ForEach limitations", "pipeline validation", "linked service authentication", or needs guidance on ADF resource limits, Set Variable restrictions, or Data Flow constraints.

This skill should be used when the user asks about "Databricks Job activity", "DatabricksJob in ADF", "ServiceNow V2 connector", "ADF managed identity", or needs guidance on Databricks serverless linked services, Snowflake connectors, or MFA enforcement.

This skill should be used when the user asks about "Fabric Lakehouse connector", "Fabric Warehouse connector", "OneLake shortcuts", "Invoke Pipeline cross-platform", or needs guidance on copying data between ADF and Microsoft Fabric workspaces.

This skill should be used when the user asks about "MSYS_NO_PATHCONV", "Git Bash path conversion", "Windows shell compatibility", or needs guidance on running ADF npm builds, PowerShell scripts, or Azure CLI from Git Bash.