{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp03_full320_20260801","cell_id":"layer_decay_and_expiration_management__astronomy_astrophysics","trajectory_id":"R","attempt_index":0,"archetype_slug":"layer_decay_and_expiration_management","domain_slug":"astronomy_astrophysics","decision":"CANDIDATE","problem_id":"superseded_astronomy_data_product_lifecycle_gap","causal_lever_id":"governed_scientific_data_product_disposition","proposal":{"problem":"HYPOTHESIS: Observatory and survey archives can accumulate superseded calibration files, pipeline intermediates, simulations, and derived data products without consistent lifecycle states. Obsolete products may remain discoverable as if authoritative and consume high-cost storage, while ad hoc deletion can destroy dependencies needed to reproduce published results or reinterpret observations.","actors_substrate":["observatory archive operators","instrument and pipeline scientists","survey data-release teams","astronomers reusing archived products","raw observations, calibrations, pipeline intermediates, simulations, catalogs, and release metadata"],"observable_state":"Multiple generations of scientific products remain active or poorly labeled; ownership, age, supersession, dependencies, retention grounds, and restorability are incomplete or inconsistent.","consequence":"Researchers may select obsolete products, archive costs and search burden grow, and cleanup may break provenance or reproducibility.","affected_objective":"Keep astronomy archives interpretable and economically bounded while preserving scientifically necessary provenance, reconstruction, and long-baseline value.","structural_mapping":[{"archetype_element":"sequential deposits","domain_realization":"Successive observing runs, pipeline executions, calibration revisions, simulations, and data releases generate versioned artifacts.","claim_kind":"INFERENCE"},{"archetype_element":"stale layers remain active","domain_realization":"Superseded calibrations or derived products may remain searchable without a prominent current/superseded state.","claim_kind":"HYPOTHESIS"},{"archetype_element":"deletion can break reconstruction","domain_realization":"Published products may depend on particular raw inputs, software contexts, calibrations, or intermediate outputs.","claim_kind":"INFERENCE"},{"archetype_element":"preservation exceptions","domain_realization":"Rare-event records, long-baseline observations, publication dependencies, proprietary commitments, and investigations may require extended retention.","claim_kind":"HYPOTHESIS"},{"archetype_element":"managed lifecycle","domain_realization":"Products receive explicit active, superseded, warm, cold, quarantined, deleted, or preserved states with recorded transitions.","claim_kind":"HYPOTHESIS"}],"component_map":[{"component":"layer_inventory_and_identity_map","status":"direct","domain_realization":"Persistent identifiers and inventory for each scientific product and version."},{"component":"deposition_order_and_age_index","status":"direct","domain_realization":"Creation time, observing epoch, pipeline run, and release sequence."},{"component":"retention_policy_matrix","status":"adapted","domain_realization":"Rules by raw data, calibration, intermediate, simulation, catalog, and release class."},{"component":"decay_function_or_aging_rule","status":"adapted","domain_realization":"Age changes access tier or review priority, but never presumes declining scientific value."},{"component":"expiration_trigger_definition","status":"direct","domain_realization":"Supersession, inactivity, elapsed project phase, or capacity thresholds trigger review."},{"component":"layer_value_and_risk_score","status":"adapted","domain_realization":"Transparent factors for reuse, uniqueness, reproducibility, misinterpretation risk, and recreation cost."},{"component":"dependency_and_reconstruction_check","status":"direct","domain_realization":"Trace publications, releases, workflows, provenance, and derived products before removal."},{"component":"pruning_or_archival_pathway","status":"direct","domain_realization":"Relabel, compact, migrate, quarantine, or delete according to class."},{"component":"preservation_exception_register","status":"direct","domain_realization":"Named holds for rare phenomena, long baselines, active research, or historical significance."},{"component":"deletion_audit_and_rollback_window","status":"direct","domain_realization":"Logged reversible quarantine precedes destruction."},{"component":"resource_budget_and_storage_tier","status":"direct","domain_realization":"Capacity and retrieval targets for hot, warm, cold, and deep archive."},{"component":"review_cadence_and_revalidation_loop","status":"direct","domain_realization":"Scheduled review of rules, holds, dependencies, and restore performance."},{"component":"legal_hold_or_compliance_override","status":"adapted","domain_realization":"Hard override for legal duties plus proprietary, investigation, and institutional commitments."},{"component":"hot_warm_cold_access_state","status":"direct","domain_realization":"Explicit access tier and expected retrieval delay per product."},{"component":"orphan_layer_detection","status":"direct","domain_realization":"Identify intermediates with no traceable live dependent or reconstruction role."},{"component":"supersession_marker","status":"direct","domain_realization":"Old product points to its successor and remains visibly non-current."},{"component":"layer_compaction_candidate_flag","status":"adapted","domain_realization":"Flag small intermediates or redundant representations for lossless packaging or deduplication."},{"component":"rehydration_or_restore_procedure","status":"direct","domain_realization":"Documented retrieval, validation, and provenance reconnection from cold storage."},{"component":"deletion_impact_estimate","status":"direct","domain_realization":"Estimate affected products, users, publications, uniqueness, and recovery cost."},{"component":"stale_context_detector","status":"adapted","domain_realization":"Detect superseded calibration, schema, pipeline context, or release authority."}],"mechanism_dispositions":[{"slug":"age_weighted_value_score","disposition":"considered_rejected","contribution_type":"NONE","adaptation_or_rejection":"Scientific value may rise with age or longer baselines; use explicit factors and age only for review priority.","counterfactual_removal":"Removal avoids false precision and does not break the lifecycle chain."},{"slug":"archive_restore_test","disposition":"selected_load_bearing","contribution_type":"SAFETY_GUARDRAIL","adaptation_or_rejection":"Sample products across formats, age bands, and tiers; validate bytes, parsing, provenance, and usability.","counterfactual_removal":"Cold migration could become unverified loss, hard-gating safe tiering."},{"slug":"cache_eviction_rule","disposition":"selected_supporting","contribution_type":"OPERATIONAL","adaptation_or_rejection":"Apply only to reproducible fast-access cache copies; eviction demotes rather than destroys canonical products.","counterfactual_removal":"The archive remains governed, but hot-tier capacity requires more manual control."},{"slug":"dependency_safe_delete_check","disposition":"selected_load_bearing","contribution_type":"SAFETY_GUARDRAIL","adaptation_or_rejection":"Trace scientific provenance and publication/release dependencies; incomplete visibility blocks destruction.","counterfactual_removal":"Cleanup could silently make results irreproducible."},{"slug":"lifecycle_storage_tiering_policy","disposition":"selected_load_bearing","contribution_type":"CORE_CAUSAL","adaptation_or_rejection":"Demote low-access products while preserving identity, metadata, and restoration paths.","counterfactual_removal":"The proposal loses its main means of reducing active-storage cost without destroying history."},{"slug":"log_rotation_and_cleanup_job","disposition":"considered_rejected","contribution_type":"NONE","adaptation_or_rejection":"Generational rotation is too content-blind for heterogeneous scientific products; reserve it for operational logs outside scope.","counterfactual_removal":"No material change to the target intervention."},{"slug":"retention_schedule","disposition":"selected_load_bearing","contribution_type":"CORE_CAUSAL","adaptation_or_rejection":"Classify scientific artifacts and encode minimums, review points, disposition paths, and named holds.","counterfactual_removal":"Disposition becomes ad hoc and exceptions cannot be governed consistently."},{"slug":"soft_delete_quarantine_window","disposition":"selected_load_bearing","contribution_type":"SAFETY_GUARDRAIL","adaptation_or_rejection":"Hide candidates while retaining recoverability; size the window by impact and retrieval cadence.","counterfactual_removal":"Early mistakes become irreversible and the bounded pilot is unsafe."},{"slug":"stale_layer_detection_dashboard","disposition":"selected_load_bearing","contribution_type":"OPERATIONAL","adaptation_or_rejection":"Expose inventory, authority, ownership, supersession, dependencies, and stale-context signals.","counterfactual_removal":"Candidates and misleading active versions remain difficult to identify."},{"slug":"time_to_live_ttl_policy","disposition":"selected_supporting","contribution_type":"OPERATIONAL","adaptation_or_rejection":"Use renewable TTLs only for declared reproducible intermediates and temporary cache copies; expiry means review or demotion, not automatic destruction.","counterfactual_removal":"More manual review is needed, but canonical-product governance survives."},{"slug":"tombstone_or_deletion_marker","disposition":"selected_load_bearing","contribution_type":"SAFETY_GUARDRAIL","adaptation_or_rejection":"Retain identifier, disposition reason, date, provenance summary, and successor or access route.","counterfactual_removal":"Missing products become ambiguous and stale references cannot resolve safely."}],"causal_chain":["Inventory and identify versions, authority, age, access, and provenance.","Detect superseded or stale-context products and classify retention obligations and exceptions.","Gate each proposed disposition on uniqueness, dependencies, reconstruction needs, and impact.","Relabel current authority, migrate low-access products, compact losslessly where justified, and quarantine deletion candidates.","Leave durable disposition markers and test restoration on schedule.","Revalidate exceptions and policies, reducing active cost and misleading discovery without sacrificing required memory."],"baseline":"Keep products on their original tier indefinitely, with occasional project-specific cleanup and inconsistent supersession labels.","nearest_rival":"Buy additional storage and improve search/version metadata while retaining every scientific product indefinitely.","authority_safety":{"affected_parties":["archive users","data-producing collaborations","authors of dependent publications","observatory operators","holders of proprietary or sensitive observing rights"],"decision_authority":"An observatory-designated archive governance group with archive operations, instrument or pipeline expertise, and scientific-data stewardship representation.","authorized_first_step":"Run a metadata-only inventory and dry-run classification on one completed, non-proprietary product family; then migrate only redundant or reproducible copies to a reversible cold tier and perform restore tests. No permanent deletion is authorized.","excluded_actions":["automatic destruction based solely on age, inactivity, score, or TTL","removal with unresolved dependencies or ownership","lossy compaction of canonical scientific measurements","overriding legal, proprietary, investigation, publication, or preservation holds","representing a superseded product as nonexistent"],"halt_rollback":"Halt on any unresolved dependency, restore failure, provenance mismatch, unexpected user demand, or disputed preservation value; restore migrated items, reinstate visibility, and retain audit records."}},"negative_tests":{"strongest_counterevidence":"A sampled archive may already have complete version authority, dependency-rich provenance, tested restoration, affordable capacity, and high reuse of old products; then lifecycle governance adds overhead without solving a material problem.","analogy_break":"Astronomical observations do not reliably decay in value: old measurements can become uniquely valuable for long time baselines or reinterpretation after new discoveries. Age may govern review or access cost, not scientific worth or deletion.","failure_condition":"The intervention fails if dependency metadata is too incomplete to distinguish disposable intermediates from unique reconstruction inputs, or if cold-tier restoration cannot meet scientific-use requirements.","problem_falsifier":"Across a representative product inventory, superseded items are already unambiguously labeled, active-tier cost and search burden are acceptable, dependencies are preserved, restore tests pass, and no unsafe cleanup backlog exists.","intervention_falsifier":"Compared with baseline on the bounded product family, the policy does not reduce active-tier footprint or stale-product selection, or it increases failed restores, broken provenance, retrieval delays beyond agreed targets, or disputed dispositions.","risks":["Irreplaceable observations or reconstruction inputs are misclassified as expendable.","Cold-tier latency impedes urgent transient or follow-up analysis.","Composite scores obscure scientific judgment.","Lossy compaction alters measurements or uncertainty information.","Incomplete external publication dependencies produce false orphan status.","Quarantined or tombstone metadata retains sensitive information longer than permitted."]},"null_rationale":null,"classification":{"candidate_kind":"MECHANISM_COMPOSITION","prior_art_status":"UNSEARCHED","evidence_maturity":"HYPOTHESIS"},"revision_change_log":{"revision_kind":"ORIGINAL","prior_problem_id":null,"prior_causal_lever_id":null,"problem_changed":false,"causal_lever_changed":false,"conceptual_changes":[],"operational_changes":[],"repairs_addressed":[]},"confidence":0.84,"generator_notes":"Closed-book structural inference from the supplied archetype and domain packet; empirical prevalence and performance claims remain hypotheses."}