| name | mcp-engine-query |
| description | Write, review, scope, execute, and analyze DAX queries in Power BI models. Use when creating or fixing DAX queries, deciding between run_query operations, investigating performance or VertiPaq issues, testing RLS access, or discovering schema before querying. |
PBI Query
Use this skill for query authoring and query analysis work. Read only the bundled references needed for the current task.
Query workflow
- Inspect schema before guessing table, column, or measure names.
- Decide whether the task is execution, performance analysis, VertiPaq inspection, or access testing.
- Keep the first query scoped and small enough to inspect safely.
- 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
Reference selection