What relationships can a generated policy dependency dataset contain?

MirrorFirm could generate a dependency graph around a policy change, rather than a folder of disconnected documents. The graph can connect a policy version to its clauses, each clause to the operational artefacts that implement it, and those artefacts to the work that uses them.

In professional services, an engagement acceptance clause can determine which declaration appears on an intake form, which approval route applies and which checks must be completed before work begins. A clause amendment may affect only one of those relationships. Treating every associated artefact as equally affected would create unnecessary work.

Useful relationship types include `implements_clause`, `supersedes`, `used_by`, `owned_by` and `depends_on`. Edges can carry reasons and validity intervals so the model can distinguish a current dependency from one that ended under an earlier process. Documents can retain section references that connect prose to the graph.

The generated package can also include deliberately missing relationships as a defined task condition. The model may need to discover a dependency from a procedure rather than receive a complete map. Buyers should specify whether graph completion, impact analysis or both form part of the intended task.