Validation Pack: AOR-13, Executable aligned to PROV-O

Unit: applied-ontology-readiness Decision: ADR-A92 item 3, rewritten on acceptance

Invariant

Executable plans and artefacts are PROV-O entities derived from their sources, without Executable importing Foundation, and without adding any functional restriction.

Test cases

ID	Given / When / Then	Level	+/-
AOR13-01	Executable spec / parsed / imports PROV-O and no Foundation document, plans and artefacts `⊑ prov:Entity`, four derivation properties `⊑ prov:wasDerivedFrom`	L1	+
AOR13-02	a compiled concept plan plus Executable’s closure / RDFS closure (`owlrl`) / the plan node and its condition are `prov:Entity`	L2	+

One command

mise run check:ontology-catalog

Pass: every tool test passes, the four in tools/test_provenance_alignment.py among them (with AOR-12).

Artefacts to inspect

ontology/mork/spec/Executable.ttl, the PROV-O alignment block.
ADR-A92 item 3 and its consequences.

Evidence for the condition on direct alignment

PROV-O (http://www.w3.org/ns/prov-o-20130430, fetched 2026-09-25) declares no owl:FunctionalProperty and no owl:InverseFunctionalProperty. Its one cardinality restriction is on prov:ActivityInfluence. It declares prov:Entity disjoint with prov:Activity, which is the one restriction the alignment passes on: a node a plan derives from cannot also be an activity.

Adversarial probe (run by the agent)

Removing exe:derivedFromEligibilityNode ⊑ prov:wasDerivedFrom failed AOR13-01 and AOR13-02.