Skip to content
AI.Tedt.orgAI Library
All Assets
Agent Skillactive

Power BI Model Review

Review a Power BI or Analysis Services semantic model (TMDL, BIM/JSON, or a pasted table/measure list) for relationship, DAX, date-handling, and naming issues, with corrected DAX for anything flagged.

Updated Jul 26, 2026Version 1.0.0

About This Asset

Power BI Model Review

Paste in a semantic model (a TMDL export, a .bim/JSON model definition, or just a written-out list of tables, relationships, and measures) and get a structural review back: relationship design, calculated-column vs. measure choices, date-table setup, naming, star-schema shape, and common DAX mistakes. Corrected DAX is included wherever a fix applies.

What it’s for

Catching the modeling mistakes that don’t show up until a report is slow or a number is subtly wrong: bi-directional relationships between two dimension tables, calculated columns doing a measure’s job, time intelligence running against a column that isn’t really a date, CALCULATE filters that quietly override the wrong thing.

What it isn’t

It cannot connect to your workspace or a live model. Everything is inferred from the definition you provide. It doesn’t measure real query time, doesn’t know your row counts, and won’t tell you a report is slow, only what in the model is likely to make it slow. Treat the output as a structured second opinion, not a performance profile.

Getting a TMDL export

Power BI Desktop’s model view (or Tabular Editor) can serialize a model to TMDL folders, or use Get-BimFile/model.bim export from an existing workspace. A pasted list of tables/relationships/ measures works too if a full export isn’t handy. The review is only as thorough as what it’s given.

Reference

Star schema guidance · Bi-directional relationship guidance · Optimizing semantic models for Copilot


Skill by Tim Karlsson (╯°□°)╯︵ ┻━┻ Works 60% of the time, every time.

Continue Exploring
Browse the Full Library

Rate and Discuss This Agent Skill

Microsoft upstream and AI.Tedt.org remain separate communities. Discovery totals combine their responses, but each discussion stays with its source.

Combined Response
▲ 00 comments

Source totals are added. They do not represent unique people.

Microsoft Upstream
▲ 00 comments
89 repository stars
AI.Tedt.org
▲ 00 comments
0 repository stars

AI.Tedt.org Discussion

Open on GitHub

Microsoft Upstream Discussion

Open on GitHub