# EM-PEO-03 provider comparison

Claude and Grok converge on a definition-only completion of reserved `WM-PER-009`, identifier-unassigned `Proficiency Scale` and `Person Capability Assertion` candidates, a Competency Assessment profile over `WM-ACT-034`, and reuse of `WM-PER-008`, `WM-XCT-017`, and the `WM-PER-013` licence profile. Both prohibit course or role inference, current competence from expired artifacts, identity merging from equal labels, and exact or reversible cross-scale equivalence.

Claude sharpened the definition-versus-holding identity split, bitemporal evidence and assessment boundaries, and the difference between concept and level mappings. Grok sharpened the exact definition/scale identities, disjoint assertion kinds, current-as-query rule, non-transitive person inference and the minimum model set. External standard details and neighboring ownership remain held unless pinned in the frozen dossier.