The XML does not show how the build flows.
A BuildGraph script is a list of nodes in XML. Reading it does not tell you which node hands files to which, what a property is worth by the time a node uses it, or which nodes end up on different agents.
Mistakes in those three areas can pass the script's validation without an error. The viewer is built to find them.
Drop a script on the page and read the report.
The viewer is a static analyser: it reads the script and never runs it. The analysis happens in your browser tab, so no file leaves your machine and no server sees it.
Drop the script
Drop a BuildGraph .xml file, or the folder it lives in, onto the page. The viewer follows every script it includes.
Read the findings
The report opens with the findings. Each one names the exact file and line.
Check the graph
A swimlane graph backs up the findings: one lane per agent, one box per node, solid wires for real data dependencies and dashed wires for ordering only.
Three things the XML hides, and the tools to follow them.
#Tags.Known limits.
Rogue BuildGraph Viewer is in beta.
It is in beta and open to try. It runs in your browser, with nothing to install, and your files stay on your machine.
There is nothing to install. Drop a BuildGraph script, or the folder it lives in, onto the page.