B05

• Input: Digital entities, activities, agents, derivations, versions, identifiers, metadata, and source conditions.
• Operation: Assign persistent identifiers and rich metadata; represent generation, attribution, derivation, specialization, and revision; retain enough inputs and context for reuse/reconstruction.
• Output: Interoperable provenance graph and reusable research objects.
• Scope conditions: Digital artifacts exchanged or reused across systems.
• Failure conditions: Identifiers are unstable; provenance is incomplete; referenced inputs are inaccessible or mutable.
• Predicted effect: Improved findability, impact tracing, interoperability, and reconstruction.
• Measurement: Retrieval/reconstruction success, lineage completeness, and affected-descendant recall.
• Implementation test: A new operator resolves IDs, traverses lineage, and reconstructs a sampled output from recorded inputs.
• Decision rights: Publisher/agent asserts provenance; consumer validates.
• Uncertainty handling: Conflicting assertions can coexist with attribution.
• Verification: Constraint/digest checks and reconstruction.
• Provenance: Core operation.
• Implementation cost: Metadata capture, storage, and governance.
• Limitations: Does not prescribe domain-specific test design. Direct: specification and primary principles paper. Inference: exact metrics.