What problem does it solve?
Refreshing Power BI semantic models reliably requires coordinating the Enhanced Refresh REST API, choosing the right refresh type, validating that data actually changed, and diagnosing failures across credentials, schemas, timeouts, and capacity limits. This Skill gives an AI agent the exact commands, scripts, and troubleshooting references to perform that workflow end to end.
Core Features & Use Cases
- Full Refresh Lifecycle: Resolve workspace and model IDs, capture a pre-refresh DAX baseline, trigger refreshes (full, automatic, dataOnly, calculate, clearValues, defragment), monitor status, cancel in-progress refreshes, and validate post-refresh data changes.
- Targeted and Large-Model Strategies: Refresh specific tables or partitions, run two-phase dataOnly + calculate refreshes, apply partialBatch commit modes, control parallelism, and work with incremental refresh policies.
- Scripted Automation: The scripts/refresh_model.py CLI wraps the Enhanced Refresh API with arguments for type, scope, commit mode, retries, polling, and cancellation.
- Troubleshooting References: Detailed guides cover credential errors, type and schema mismatches, timeouts, incremental refresh issues, capacity throttling, and per-table failure isolation.
- Use Case: After an ETL pipeline loads new rows into a Fabric lakehouse, ask the agent to refresh the FactSales table, poll until completion, and verify via DAX that the row count and max date increased.
Quick Start
Ask the agent to refresh the semantic model in your workspace and confirm that new data arrived by comparing row counts before and after the refresh.