PBI Query

Use this skill for query authoring and query analysis work. Read only the bundled references needed for the current task.

Query workflow

1. Inspect schema before guessing table, column, or measure names.
2. Decide whether the task is execution, performance analysis, VertiPaq inspection, or access testing.
3. Keep the first query scoped and small enough to inspect safely.
4. Expand only after the initial query shape is correct.

Choose the operation

• Use execute when the user needs query results or a validation query.
• Use analyze when the user needs timings, operator-level diagnosis, or performance findings.
• Use vertipaq when the user needs storage footprint, table size, or cardinality clues.
• Use test_access when the user needs role-by-role access validation.

Author safely

• Read dax-query-guide before writing new or corrected query text.
• Prefer schema discovery with list_model before inventing names or relationships.
• Keep the default query scoped with filters, TOPN, or targeted grouping.
• Add ordering for multi-row outputs so results are stable and readable.
• Treat nulls in result rows as query semantics until the model proves otherwise.

Diagnose performance

• Read query-performance-guide before interpreting timings or trace output.
• Read dax-query-plan-reference when a plan contains unfamiliar operators.
• Read vertipaq-optimization-guide when the issue looks like storage bloat or high-cardinality pressure.
• Read performance-remediation-playbook when the user wants a remediation sequence instead of isolated findings.

Reference selection

• Read dax-query-guide first for syntax, quoting rules, and DAX query structure.
• Read query-performance-guide for timings, runs, and performance interpretation.
• Read dax-query-plan-reference for plan operator definitions.
• Read vertipaq-optimization-guide for storage optimization guidance.
• Read performance-remediation-playbook for end-to-end tuning workflow.