Measure Chain

Measure Chain turns a semantic model’s DAX into a dependency graph. Paste TMDL or drop .tmdl files in, and it shows which measures feed which, grouped by display folder. I built it first as a Fabric Data App, then as a single HTML file, with a sibling Query Chain that does the same for Power Query M.

##How it works

• Measure Lineage draws the folder-grouped graph with a barycenter layout. Click a measure for its DAX and both reference lists, or use focus mode to narrow the canvas to one measure and its direct neighbours.
• Model Metadata lists tables, columns and measures with every property the TMDL declared. DAX Breakdown re-indents one measure with the measures it references nested underneath.
• The HTML edition is one 345 KB file: no framework, no network, no sign-in. Nothing leaves the tab.
• The parsers exist twice: JavaScript for the browser, and stdlib-only Python with a CLI so an agent skill can call them. Tests assert the two agree byte for byte (139 checks for TMDL, 112 for M, plus 24 parity dumps).

##Why it matters

Two implementations that can disagree are one implementation and one liability. Testing them against each other means the page and the CLI can never disagree about what a model contains.