14. Open Questions

(Informative)

1. Naming. The domain bpmstack.org has been registered, establishing the public identity for this specification. The name “BPM/Agent Stack” communicates the bridge between the source discipline and the application domain. Whether this name optimally serves both the BPM community (which recognizes BPM) and the agent community (which does not) remains an open question for market feedback.

2. Specification scope. Should this specify a complete process modeling notation for agent execution, or a governance attribute model and structural pattern set that sits on top of existing standards? This version takes the latter approach.

3. CMMN integration. Anthropic’s distinction between “workflows” (predefined paths) and “agents” (dynamic, model-directed) maps to the BPM distinction between structured processes (BPMN) and adaptive case management (CMMN). The current specification focuses on BPMN-type structured processes. CMMN-type adaptive patterns merit separate treatment.

4. Five stitching intersections. Does each of the five Intent Primitives have an independent stitching intersection with the BPM/Agent Stack, or is Key Tasks the unique structural joint with the other four primitives flowing through it? (§7.5) The reclassification of execution governance concerns to this specification (v1.1) does not change this structural question — the stitching point remains at the boundary between governance context and execution governance regardless of which specification document those concerns reside in.

5. Mathematical foundations. Does the BPM/Agent Stack need its own formal mathematical grounding (parallel to the Intent Stack’s fiber bundle structures), or is its authority properly derived from the existing formalization of BPMN 2.0, DMN 1.0, and the BPM CBOK?

6. Adjacency as discovery mechanism. The visualization analysis that informed this specification revealed emergent structural insights through spatial adjacency of previously separate concepts. Is adjacency creation — through visualization, cross-domain concept mapping, or structured juxtaposition — a formalizable governance practice? Is it an operational mechanism for Intent Stack L4 (Intent Discovery)?

7. Context/Memory/Intent formalization. The disambiguation in Clause 11 is preliminary. A formal characterization of the structural relationships between context, memory, and intent — parallel to the Intent Stack’s formal characterization of Intent itself — would strengthen the three-layer architecture’s conceptual foundations.

8. Integration protocol. Formal specification needed for: how the Key Tasks primitive interfaces with BPM/Agent Stack process entry at Intent Stack L2 (Specification), and how the BPM/Agent Stack provides execution evidence back to Intent Stack L1 (Runtime Alignment). The stitching mechanism is specified narratively in Clause 7; a formal protocol specification would enable implementation.