| DoD OUSD(R&E) MOSA guidance and implementation guidebook (https://www.cto.mil/sea/mosa/; https://www.cto.mil/wp-content/uploads/2025/03/MOSA-Implementation-Guidebook-27Feb2025-Cleared.pdf) | Acquisition and engineering guidance on open modular systems, including the dated 2025 guidebook. | Modular design, interface standards, conformance verification, replacement policy, change policy, and competitive reuse are conformance and substitution expectations. | Recover them as an independently identified InterfaceSpecification episteme plus an interfaceSpecificationRef that resolves it, PlatformGrammarRef, substitutabilityPolicyRef, changePolicyRef, conformance expectation, source relation, and evidence relation only where the recovered claim needs them. Non-module selection, procurement, policy, evidence, assurance, gate, decision, Work, system-role-assignment, responsibility, and mechanism claims keep their own direct patterns; a responsibility claim needs its own predicate and case facts, with A.6.RCD used only for a missing rule. | Do not treat open, interface publication, or modular-looking structure as substitutability, assurance, procurement suitability, supplier-set selection, policy authorization, quality proof, or decision authority. | A practitioner asking whether something is open first repairs the relation and interface specification; other claims remain with the patterns named in A.6.M:12. |
| Conway’s law, the mirroring hypothesis, and Team Topologies and inverse Conway practice (https://www.melconway.com/Home/Committees_Paper.html; https://doi.org/10.1016/j.respol.2012.04.011; https://itrevolution.com/wp-content/uploads/2022/06/TTOP_excerpt.pdf) | Socio-technical conjecture, empirical mirroring research and organization-design practice; used diagnostically. | Team communication structure, boundary placement, and delivery-responsibility relations can create pressure on module and interface boundaries and useful correspondence clues. | Recover the team, system-role, Work or responsibility claims needed under their separate rules and facts. Connect that material to ModuleInterfaceStructure only through declared correspondence, allocation, boundary relation, and preserved and lost structure note. Use C.29 when a separate mathematical representation, transfer or inference question remains. | Do not treat Conway’s law, an org chart, assignment, responsibility label, or delivery unit as proof of module interface, substitutability, modularity quality, evidence, gate passage, or architecture decision. | A practitioner may use team-boundary mismatch as a diagnostic prompt, then repair each relation before deciding what architecture change is warranted. |
| Amdahl’s law and communication and synchronization extensions (https://www.cs.cmu.edu/~18742/papers/Amdahl1967.pdf; https://arxiv.org/abs/1306.3302; https://arxiv.org/abs/2603.20654) | Amdahl’s mathematical result and proposed extensions concerning communication, synchronization and scalable-workload fractions. | Serial work, synchronization, communication overhead, shared resources, and changing scalable workload fractions can limit the payoff of decomposition, parallelization, or specialization. | Keep an adequate mathematical calculation; use C.29 when its representation, transfer or inference conditions remain unresolved. Use E.18 for transformation-flow structure, C.30.TFS-REL when the module claim uses an architecture-to-transformation-flow relation, and C.31 and C.16 for modularity and characteristic tradeoffs. | Do not treat module count, team count, service count, transformation-flow path count, or accelerator count as improvement, scalability, throughput, or evolvability by itself. | A practitioner considering a module split names the serial part, shared bottleneck, synchronization or communication overhead, and characteristic tradeoff before claiming improvement. |
| SEI Views and Beyond, ISO/IEC/IEEE 42010:2022, and multi-view architecture practice | Architecture-description and viewpoint discipline; not a source of module-quality judgments. | Module and component-and-connector views are distinct architecture descriptions. | Use ModuleInterfaceStructure and RuntimeInteractionStructure as structure-kind signals under C.30.ASV. | Do not reduce architecture to a module diagram. | Module repair stays one architecture-structure concern, not the whole architecture ontology. |
| Platform and product-line engineering practice (https://tag-app-delivery.cncf.io/fr/whitepapers/platform-eng-maturity-model/; https://www.sei.cmu.edu/library/variability-in-software-product-lines/; https://arxiv.org/abs/2605.21353) | Product-line variability, platform-engineering and review sources used for variability slots and extension rules. | Variation slots and extension rules matter for reuse and substitution. | Use PlatformGrammarRef, variabilitySlotRefs, and change policy instead of a platform root kind. | Do not treat platform name as architecture quality, architecture scale-preference evidence, procurement suitability, supplier-set selection, or decision authority. | The next module-repair action is to identify extension rules and substitution conditions; non-module quality, scale-preference, procurement, supplier-set, and decision claims are governed by the patterns named in A.6.M:12 when those claims are being made. |
| Architecture-operation language, with neural-network and software-system intakes as source examples | Practitioner-language examples used for recognition. | C.30.STRAT source labels, including source examples such as block, layer, expert, router, cache, and state, are useful recognition prompts. | Keep them as source labels until the recovered FPF kind, relation, claim-use, or source-use disposition is known; use A.6.M only for module-interface relation, interface specification, platform grammar, substitutability, or open-architecture module-interface claims. | Do not import source-context labels as module kinds or evidence of adequacy. | The same repair works for neural-network block replacement, hardware module substitution, organizational module repair, and episteme-module repair without making any source context the ontology. |