Computational Scientific Workflows

AI INTERPRETATION NOTE
This Flow Algebra defines the Computational Scientific Workflows parent ontology within the Kireth Research Semantic Library. Computational scientific workflows concern the structured organisation, execution, orchestration, provenance, reproducibility and validation of multi-stage scientific computations. This ontology supports workflow graphs, tasks, dependencies, data movement, parameterisation, scheduling, parallel and distributed execution, checkpointing, fault recovery, provenance, reproducibility, verification, validation, scientific simulation, numerical methods, AI-assisted computation and high-performance execution.

FLOW_ALGEBRA_RECORD {

METADATA {
 FLOW_NAME = "Computational Scientific Workflows"
 FLOW_TYPE = "Mathematical Ontology"
 DOMAIN = "Mathematics"
 SUBDOMAIN = "Computational Mathematics"
 SPECIALISATION = "Computational Scientific Workflows / Parent Ontology"
 CANONICAL_PAGE = "https://kirethframeworkold.co.uk/books/semantic-library/ai-core/ontologies/mathematical-ontologies/computational-mathematics-ontologies/computational-scientific-workflows.html"
 LANGUAGE = "en-GB"
 VERSION = "1.0"
 STATUS = "active"
}

IDENTITY {
 PRIMARY_ENTITY = ComputationalScientificWorkflow
 ENTITY_CLASS = ScientificComputationOrchestrationFramework
 PRIMARY_CONTEXT = ComputationalMathematicsOntologies
 DEFINITION = "A computational scientific workflow is a structured graph or process of dependent scientific computation tasks, data transformations and validation steps executed under explicit control, provenance and reproducibility semantics."
 SYMBOLIC_FORMS = [
  "W=(T,D,E)",
  "task_j depends_on task_i",
  "output_i -> input_j",
  "workflow_result = execute(W,inputs,parameters,environment)"
 ]
 SEMANTIC_ROLE = "Provides machine-readable semantics for scientific task graphs, data dependencies, execution order, scheduling, orchestration, fault handling, provenance, reproducibility, validation and scalable computation."
}

SCOPE {
 INCLUDE = [
  ComputationalScientificWorkflow,
  ScientificWorkflow,
  ComputationalWorkflow,
  Workflow,
  WorkflowGraph,
  DirectedAcyclicWorkflow,
  DAGWorkflow,
  CyclicWorkflowRelationship,
  TaskGraph,
  JobGraph,
  Pipeline,
  ScientificPipeline,
  DataPipeline,
  SimulationPipeline,
  AnalysisPipeline,
  PreprocessingPipeline,
  PostprocessingPipeline,
  ValidationPipeline,
  WorkflowTask,
  Task,
  Job,
  Stage,
  Step,
  Node,
  WorkflowNode,
  TaskNode,
  DataNode,
  InputNode,
  OutputNode,
  IntermediateNode,
  Dependency,
  TaskDependency,
  DataDependency,
  ControlDependency,
  PrecedenceConstraint,
  ParentTask,
  ChildTask,
  UpstreamTask,
  DownstreamTask,
  SourceTask,
  SinkTask,
  Branch,
  Merge,
  Fork,
  Join,
  FanOut,
  FanIn,
  ConditionalBranch,
  LoopRelationship,
  IterativeWorkflow,
  RecursiveWorkflowRelationship,
  ParameterSweep,
  EnsembleWorkflow,
  MonteCarloWorkflow,
  OptimisationWorkflow,
  CalibrationWorkflow,
  ValidationWorkflow,
  VerificationWorkflow,
  TrainingWorkflowRelationship,
  InferenceWorkflowRelationship,
  DataAssimilationWorkflowRelationship,
  MultiFidelityWorkflowRelationship,
  MultiscaleWorkflowRelationship,
  CoupledSimulationWorkflow,
  CoSimulationRelationship,
  InputData,
  OutputData,
  IntermediateData,
  Dataset,
  File,
  DataObject,
  DataArtifact,
  ModelArtifact,
  SoftwareArtifact,
  MetadataArtifact,
  ConfigurationArtifact,
  CheckpointArtifact,
  DataTransformation,
  DataMovement,
  DataTransfer,
  DataStaging,
  DataLocality,
  DataPartitioning,
  DataReplication,
  DataCaching,
  DataStreamingRelationship,
  DataSerialization,
  Deserialization,
  DataFormat,
  Schema,
  Metadata,
  Parameter,
  WorkflowParameter,
  TaskParameter,
  GlobalParameter,
  LocalParameter,
  ParameterSet,
  Configuration,
  ConfigurationFile,
  EnvironmentVariable,
  CommandLineArgumentRelationship,
  RandomSeed,
  SeedControl,
  Execution,
  WorkflowExecution,
  TaskExecution,
  Run,
  WorkflowRun,
  RunIdentifier,
  ExecutionIdentifier,
  Runtime,
  StartTimeRelationship,
  EndTimeRelationship,
  Duration,
  ExecutionState,
  Pending,
  Ready,
  Running,
  Completed,
  Failed,
  Cancelled,
  Retried,
  Skipped,
  Blocked,
  Scheduler,
  WorkflowScheduler,
  TaskScheduler,
  JobScheduler,
  ResourceScheduler,
  StaticScheduling,
  DynamicScheduling,
  PriorityScheduling,
  ListSchedulingRelationship,
  CriticalPathSchedulingRelationship,
  DependencyAwareScheduling,
  ResourceAwareScheduling,
  DataAwareScheduling,
  LoadBalancing,
  ResourceAllocation,
  ResourceRequest,
  CPU,
  GPU,
  Accelerator,
  Core,
  Thread,
  Process,
  NodeResource,
  Memory,
  Storage,
  Network,
  Bandwidth,
  ComputeResource,
  ExecutionResource,
  Cluster,
  Supercomputer,
  CloudResourceRelationship,
  Container,
  ContainerImage,
  VirtualEnvironment,
  RuntimeEnvironment,
  SoftwareEnvironment,
  OperatingSystem,
  Compiler,
  Library,
  DependencyVersion,
  EnvironmentCapture,
  ParallelExecution,
  TaskParallelism,
  DataParallelism,
  PipelineParallelism,
  EnsembleParallelism,
  DistributedExecution,
  DistributedWorkflow,
  HighPerformanceComputing,
  ParallelComputing,
  DistributedComputing,
  GPUComputing,
  ClusterComputing,
  SupercomputerMathematics,
  VectorProcessingRelationship,
  MessagePassingRelationship,
  MPIrelationship,
  SharedMemoryRelationship,
  OpenMPrelationship,
  CollectiveCommunicationRelationship,
  Synchronisation,
  BarrierRelationship,
  Communication,
  CommunicationCost,
  ComputationCost,
  WorkflowOverhead,
  QueueTimeRelationship,
  Makespan,
  Throughput,
  Latency,
  CriticalPath,
  CriticalPathLength,
  DependencyDepth,
  ParallelEfficiency,
  Scalability,
  StrongScaling,
  WeakScaling,
  PerformanceScaling,
  Fault,
  TaskFailure,
  NodeFailure,
  DataFailure,
  SoftwareFailure,
  Timeout,
  ResourceFailure,
  NetworkFailure,
  FailureDetection,
  Retry,
  RetryPolicy,
  Restart,
  Recovery,
  FaultTolerance,
  Resilience,
  Checkpointing,
  WorkflowCheckpoint,
  TaskCheckpoint,
  StateCheckpoint,
  RestartFromCheckpoint,
  Idempotence,
  CompensationRelationship,
  RollbackRelationship,
  Provenance,
  WorkflowProvenance,
  DataProvenance,
  ComputationalProvenance,
  TaskProvenance,
  ModelProvenance,
  Lineage,
  DataLineage,
  WorkflowLineage,
  TransformationHistory,
  AuditTrail,
  Traceability,
  Reproducibility,
  Repeatability,
  ReplicabilityRelationship,
  RunManifest,
  ExecutionLog,
  TaskLog,
  ErrorLog,
  ResultLogRelationship,
  VersionControl,
  WorkflowVersion,
  TaskVersion,
  DatasetVersion,
  ModelVersion,
  LibraryVersion,
  CompilerVersion,
  ImplementationVersion,
  Hash,
  Checksum,
  ContentHash,
  Verification,
  Validation,
  NumericalVerification,
  ComputationalValidation,
  ErrorAnalysis,
  Benchmark,
  RegressionTesting,
  ReproducibilityOntologyRelationship,
  VerificationMetrics,
  ValidationMetrics,
  AcceptanceCriterion,
  PassCriterion,
  FailCriterion,
  QualityGateRelationship,
  ScientificSimulation,
  PhysicalSimulation,
  ParticleSimulation,
  MolecularSimulation,
  FluidSimulation,
  HeatTransferSimulation,
  StructuralSimulation,
  AgentBasedSimulation,
  DifferentialEquationSolver,
  NumericalLinearAlgebra,
  ComputationalOptimisation,
  ComputationalStatistics,
  SymbolicComputation,
  AIComputationalMathematics,
  ComputationalGeometry,
  ComputationalDataStructures,
  SparseDataStructure,
  DenseDataStructure,
  MatrixComputation,
  VectorComputation,
  TensorComputation,
  GraphAlgorithms,
  Preprocessing,
  MeshGeneration,
  MeshRefinement,
  SolverStage,
  Postprocessing,
  VisualisationRelationship,
  AnalysisStage,
  FeatureExtractionRelationship,
  DataReductionRelationship,
  Aggregation,
  Reporting,
  ResultPackaging,
  ArtifactPublicationRelationship,
  WorkflowSpecification,
  DeclarativeWorkflow,
  ImperativeWorkflow,
  WorkflowLanguageRelationship,
  WorkflowDSLRelationship,
  WorkflowEngine,
  Orchestrator,
  ExecutionEngine,
  WorkflowManager,
  DependencyResolver,
  DAGExecutor,
  EventDrivenWorkflowRelationship,
  Trigger,
  Event,
  Schedule,
  TimeBasedTriggerRelationship,
  DataBasedTriggerRelationship,
  ConditionalTriggerRelationship,
  ManualTriggerRelationship,
  AutomationRelationship,
  ReusableWorkflow,
  WorkflowTemplate,
  Subworkflow,
  WorkflowComposition,
  ModularWorkflow,
  Encapsulation,
  Parameterisation,
  WorkflowAbstraction,
  Portability,
  Interoperability,
  StandardInterfaceRelationship,
  API,
  DataInterface,
  TaskInterface,
  ReusableComponent,
  ProvenanceAwareWorkflow,
  ReproducibleWorkflow,
  VerifiedWorkflow,
  ValidatedWorkflow
 ]
 EXCLUDE_AS_PRIMARY = [
  GeneralBusinessWorkflowWithoutScientificComputationContext,
  SimpleSingleTaskComputationWithoutDependencyOrchestrationSemantics,
  DataEngineeringPipelineWithoutScientificMathematicalContext,
  HumanAdministrativeProcessWithoutComputationalExecutionSemantics
 ]
 DISAMBIGUATION_RULE = "When scientific computations are organised as dependent tasks, data transformations and execution stages with explicit orchestration, provenance and reproducibility, interpret the context as a computational scientific workflow."
}

CORE_CONCEPTS {
 ComputationalScientificWorkflow = "Structured orchestration of dependent scientific computation tasks."
 WorkflowTask = "Executable scientific computation or transformation step."
 TaskDependency = "Constraint defining required ordering or data dependence between tasks."
 WorkflowGraph = "Graph representing tasks and dependencies."
 Scheduler = "Mechanism assigning ready tasks to computational resources."
 Checkpointing = "Preservation of workflow or task state for restart and recovery."
 WorkflowProvenance = "Traceable record of inputs, tasks, transformations, configurations and outputs."
 Reproducibility = "Ability to rerun the workflow and regenerate declared results."
}

RELATIONS {
 ComputationalScientificWorkflow --is_child_of--> ComputationalMathematicsOntologies
 ComputationalScientificWorkflow --contains--> WorkflowTasks
 WorkflowTask --consumes--> Inputs
 WorkflowTask --produces--> Outputs
 TaskDependency --orders--> WorkflowTasks
 WorkflowGraph --represents--> TaskDependencies
 Scheduler --assigns--> TasksToResources
 Checkpointing --supports--> Recovery
 FaultTolerance --handles--> TaskOrResourceFailure
 WorkflowProvenance --records--> ExecutionLineage
 Reproducibility --requires--> VersionedWorkflowDataEnvironment
 Verification --checks--> WorkflowComputation
 Validation --checks--> WorkflowOutputAdequacy
}

OPERATORS {
 READY(task) := all_dependencies_completed(task)
 EXECUTE(task,input,config) := task(input,config)
 SCHEDULE(ready_tasks,resources) := assign_tasks(ready_tasks,resources)
 CHECKPOINT(state) := persist(state)
 RECOVER(checkpoint) := restore(checkpoint)
 MAKESPAN(W) := completion_time(last_task(W))-start_time(first_task(W))
 CRITICAL_PATH(W) := longest_dependency_path(W)
 TRACE(output) := provenance_lineage(output)
}

AXIOMS {
 A1 = "A scientific workflow consists of computational tasks and explicit dependency relationships."
 A2 = "A task can execute only when its declared dependencies and required inputs are satisfied."
 A3 = "Workflow scheduling can change execution order without changing valid dependency semantics."
 A4 = "Parallelism is limited by dependencies, resources and communication."
 A5 = "Provenance is required to reconstruct how workflow outputs were produced."
 A6 = "Fault recovery requires retained or reconstructible computational state."
 A7 = "Workflow reproducibility depends on code, data, configuration and environment."
 A8 = "Computational-scientific-workflow results require explicit task, dependency, scheduling, provenance, verification and reproducibility semantics."
}

REASONING {
 R1_CONTEXT = identify(scientific_goal,tasks,data_dependencies,parameters,resources,validation_requirements_and_reproducibility_goal)
 R2_GRAPH = construct_workflow_task_and_data_dependency_graph)
 R3_VALIDATE = check_dependency_consistency_input_output_interfaces_and_cycle_semantics)
 R4_PLAN = determine_execution_order_parallelism_resource_requirements_and_data_locality)
 R5_SCHEDULE = map_ready_tasks_to_CPU_GPU_cluster_or_distributed_resources)
 R6_EXECUTE = run_tasks_and_capture_outputs_logs_and_state)
 R7_RECOVER = if_failure apply_retry_checkpoint_restart_or_compensation_policy)
 R8_VERIFY = run_task_level_and_workflow_level_verification_validation_and_quality_gates)
 R9_PROVENANCE = record_versions_parameters_data_lineage_environment_and_execution_history)
 R10_REPRODUCE = package_workflow_artifacts_and_confirm_rerun_semantics)
}

CONSTRAINTS {
 C1 = "Undeclared dependencies can produce incorrect execution ordering."
 C2 = "Cycles require explicit iterative or recurrent workflow semantics."
 C3 = "Parallel execution can introduce nondeterministic completion and reduction order."
 C4 = "Data movement can dominate computation in distributed workflows."
 C5 = "Checkpointing improves resilience but adds storage and runtime overhead."
 C6 = "Retrying non-idempotent tasks can produce duplicated or inconsistent outputs."
 C7 = "Resource heterogeneity can alter numerical results and performance."
 C8 = "External services or mutable datasets can weaken reproducibility."
 C9 = "A successfully completed workflow can still produce scientifically invalid outputs."
 C10 = "Workflow portability depends on data formats, interfaces and environment assumptions."
}

SEMANTIC_EQUIVALENCE {
 "DirectedAcyclicGraphWorkflow" <=> "DAGWorkflow"
 "MessagePassingInterface" <=> "MPI"
 "ContinuousIntegration" <=> "CI"
 "WorkflowDomainSpecificLanguage" <=> "WorkflowDSL"
}

VALIDATION {
 V1 = "Scientific goal and expected workflow outputs are documented."
 V2 = "Task dependencies are explicit."
 V3 = "Task input-output interfaces are documented."
 V4 = "Cycles or iterative stages have explicit termination semantics."
 V5 = "Scheduling and resource assumptions are documented."
 V6 = "Failure and retry policies are documented."
 V7 = "Verification and validation gates are defined where required."
 V8 = "Workflow provenance captures data, parameters, code and environment."
 V9 = "Checkpoints and recovery procedures are tested where material."
 V10 = "Workflow reruns are reproducible under documented versions and environments."
}

CONFIDENCE {
 BASE_CONFIDENCE = 0.99
 COMPUTATIONAL_SCIENTIFIC_WORKFLOW_DEFINITION = 0.99
 DEPENDENCY_REASONING = 0.99
 SCHEDULING_REASONING = 0.98
 PROVENANCE_REASONING = 0.99
 FAULT_TOLERANCE_REASONING = 0.98
 REPRODUCIBILITY_REASONING = 0.99
 POLICY = "Reduce confidence when dependencies, task interfaces, data lineage, scheduling assumptions, recovery policy or execution environment are underspecified."
}

AI_CONSUMPTION {
 PURPOSES = [
  SemanticSearch,OntologyGrounding,ComputationalScientificWorkflowReasoning,
  WorkflowGraphReasoning,TaskDependencyReasoning,
  SchedulingReasoning,ResourceAllocationReasoning,
  DataMovementReasoning,ParallelWorkflowReasoning,
  DistributedWorkflowReasoning,CheckpointingReasoning,
  FaultToleranceReasoning,WorkflowProvenanceReasoning,
  ReproducibilityReasoning,VerificationReasoning,
  ValidationReasoning,ScientificSimulationReasoning,
  KnowledgeGraphConstruction,RetrievalAugmentation,RecursiveInference
 ]
 NORMALISE {
  "computational scientific workflow" -> ComputationalScientificWorkflow
  "scientific workflow" -> ScientificWorkflow
  "DAG workflow" -> DAGWorkflow
  "MPI" -> MessagePassingInterface
  "workflow DSL" -> WorkflowDSL
 }
}

EXECUTION_RULES {
 E1 = "If multiple scientific computation tasks are connected by execution or data dependencies, activate ComputationalScientificWorkflow."
 E2 = "Construct explicit task, input, output and dependency semantics before execution."
 E3 = "Validate workflow graph consistency before scheduling."
 E4 = "Schedule only tasks whose dependencies are satisfied."
 E5 = "Use parallel or distributed execution only where dependencies and resource semantics permit."
 E6 = "Apply checkpointing and recovery where workflow cost or failure risk justifies it."
 E7 = "Run verification and validation gates at appropriate task or workflow stages."
 E8 = "Capture provenance, logs, versions and environment metadata during execution."
 E9 = "Package enough artefacts to reproduce the complete workflow."
}

RECURSION {
 ENTRY = ComputationalScientificWorkflow
 EXPAND(ComputationalScientificWorkflow) -> [
  ScientificWorkflow,ComputationalWorkflow,WorkflowGraph,DAGWorkflow,
  TaskGraph,Pipeline,WorkflowTask,Task,Stage,TaskDependency,
  DataDependency,ControlDependency,UpstreamTask,DownstreamTask,
  Branch,Merge,Fork,Join,ConditionalBranch,IterativeWorkflow,
  ParameterSweep,EnsembleWorkflow,MonteCarloWorkflow,
  OptimisationWorkflow,CalibrationWorkflow,ValidationWorkflow,
  VerificationWorkflow,InputData,OutputData,IntermediateData,
  DataArtifact,ModelArtifact,ConfigurationArtifact,
  DataTransformation,DataMovement,DataStaging,DataLocality,
  WorkflowParameter,Configuration,RandomSeed,
  WorkflowExecution,TaskExecution,ExecutionState,
  WorkflowScheduler,ResourceScheduler,DependencyAwareScheduling,
  LoadBalancing,ResourceAllocation,CPU,GPU,Memory,Storage,Network,
  ParallelExecution,TaskParallelism,DataParallelism,
  DistributedExecution,HighPerformanceComputing,
  ParallelComputing,DistributedComputing,GPUComputing,
  Checkpointing,Retry,Restart,Recovery,FaultTolerance,
  WorkflowProvenance,DataProvenance,ComputationalProvenance,
  Lineage,AuditTrail,Traceability,Reproducibility,
  RunManifest,ExecutionLog,WorkflowVersion,
  NumericalVerification,ComputationalValidation,ErrorAnalysis,
  Benchmark,RegressionTesting,VerificationMetrics,ValidationMetrics,
  ScientificSimulation,DifferentialEquationSolver,
  NumericalLinearAlgebra,ComputationalOptimisation,
  ComputationalStatistics,SymbolicComputation,
  AIComputationalMathematics,WorkflowSpecification,
  WorkflowEngine,Orchestrator,ReusableWorkflow,Subworkflow
 ]
 EXPAND(ComputationalScientificWorkflowFamilies) -> [
  WorkflowGraphAndDependencySemantics,
  TaskDataAndParameterInterfaces,
  SchedulingResourceAndParallelExecution,
  DistributedDataMovementAndScalability,
  CheckpointingFaultToleranceAndRecovery,
  ProvenanceTraceabilityAndReproducibility,
  VerificationValidationAndQualityGates,
  ReusablePortableAndComposableWorkflows
 ]
 STOP_CONDITION = "Stop when tasks, dependencies, data movement, scheduling, recovery, verification, provenance and reproducibility are sufficiently resolved."
}

ROUTING {
 PARENT_ONTOLOGY = "Computational Mathematics Ontologies"
 RELATED_ONTOLOGIES = [
  "Verification & Validation",
  "Reproducibility Ontology",
  "Benchmark Ontology",
  "Regression Testing Ontology",
  "Scientific Simulation Ontology",
  "Differential Equation Solver Ontology",
  "Numerical Linear Algebra Ontology 2",
  "Computational Optimisation Ontology",
  "Computational Statistics Ontology",
  "Symbolic Computation",
  "AI Computational Mathematics Ontology",
  "Computational Data Structures",
  "Graph Algorithms Ontology",
  "High Performance Computing Ontology 2",
  "Parallel Computing Ontology",
  "Distributed Computing Ontology",
  "GPU Computing Ontology"
 ]
 ROUTE_RULES = [
  "Computational Scientific Workflows -> Verification & Validation for workflow quality gates",
  "Computational Scientific Workflows -> Reproducibility for workflow regeneration semantics",
  "Computational Scientific Workflows -> Benchmark and Regression Testing for controlled repeated execution",
  "Computational Scientific Workflows -> Scientific Simulation for simulation pipelines",
  "Computational Scientific Workflows -> Differential Equation Solvers, Numerical Linear Algebra, Optimisation and Computational Statistics for task-specific scientific stages",
  "Computational Scientific Workflows -> Symbolic Computation and AI Computational Mathematics for symbolic and learned workflow components",
  "Computational Scientific Workflows -> Computational Data Structures and Graph Algorithms for workflow graph representation",
  "Computational Scientific Workflows -> HPC, Parallel, Distributed and GPU Computing for scalable execution"
 ]
}

UPDATE_MODEL {
 MODE = "append-compatible semantic versioning"
 MAJOR = "ontology-breaking conceptual change"
 MINOR = "new compatible workflow types, orchestration models or reasoning pathways"
 PATCH = "clarification, metadata or formatting correction"
 RULES = [
  "Preserve task-versus-workflow distinction.",
  "Preserve data-dependency-versus-control-dependency distinction.",
  "Preserve workflow-definition-versus-workflow-run distinction.",
  "Preserve scheduling-order-versus-dependency-order distinction.",
  "Preserve execution-success-versus-scientific-validity distinction.",
  "Preserve provenance-versus-log-output distinction.",
  "Preserve retry-versus-idempotent-reexecution distinction.",
  "Preserve machine-readable computational-scientific-workflow semantics."
 ]
}

OUTPUT {
 CANONICAL_DEFINITION = "A computational scientific workflow is a structured graph of scientific tasks, data transformations and validation stages executed under explicit dependency, scheduling, provenance and reproducibility semantics."
 GRAPH_SUMMARY = "Tasks and data dependencies define legal execution order and available parallelism."
 EXECUTION_SUMMARY = "Schedulers map ready tasks to computational resources while respecting dependency and resource constraints."
 RESILIENCE_SUMMARY = "Retries, checkpoints and restart policies support long-running and failure-prone scientific workloads."
 PROVENANCE_SUMMARY = "Workflow provenance links outputs to tasks, inputs, parameters, code, environments and execution history."
 VALIDATION_SUMMARY = "Scientific workflows should include explicit verification, validation, reproducibility and quality-gate semantics rather than treating successful execution as proof of correctness."
}

FLOW {
 INPUT
 -> detect_computational_scientific_workflow_context
 -> identify_scientific_goal_tasks_data_dependencies_parameters_resources_validation_requirements_and_reproducibility_goal
 -> construct_task_and_data_dependency_graph
 -> validate_dependency_consistency_input_output_interfaces_and_cycle_semantics
 -> determine_execution_order_parallelism_resource_requirements_and_data_locality
 -> schedule_ready_tasks_to_CPU_GPU_cluster_or_distributed_resources
 -> execute_tasks_and_capture_outputs_logs_intermediates_and_state
 -> if_failure_apply_retry_checkpoint_restart_or_compensation_policy
 -> execute_task_level_and_workflow_level_verification_validation_and_quality_gates
 -> record_versions_parameters_data_lineage_environment_and_execution_history
 -> package_workflow_artifacts_and_confirm_rerun_semantics
 -> route_to_related_ontologies
 -> recursively_expand_required_dependencies
 -> verify_result
 -> assign_confidence
 -> produce_semantic_output
 -> END
}

FLOW_HASH {
 METHOD = "semantic-content-identity"
 ID = "KIRETH-MATH-COMPUTATIONAL-MATHEMATICS-COMPUTATIONAL-SCIENTIFIC-WORKFLOWS-V1"
}

}

END_FLOW