{"cases":[{"alternative_route_blocks":[],"case_id":"E14A039__DIRECT","case_type":"DIRECT_POSITIVE","domain":"Beverage manufacturing operations","intended_route_id":"B1","omitted_condition_id":null,"remedy_leakage_audit":"The statement reports operational symptoms and consequences without proposing equipment, policy, staffing, or process changes.","route_evidence":[{"condition_id":"ext001_constraint_envelope_adjustment_a01","intended_status":"SATISFIED","scenario_evidence":"Existing production rules prohibit timely adjustments to cleaning sequences and line assignments, leaving usable capacity idle and preventing effective responses to changing contamination risk."}],"scenario_text":"A regional beverage plant produces dairy-based drinks, fruit beverages, and bottled water on four shared filling lines. Its operating manual assigns each product family a fixed line and permits only one complete sanitation cycle per line during the overnight shutdown. These rules were written when the plant handled half its current assortment. Demand now changes sharply during the week, but supervisors cannot move a short dairy run onto an available compatible line or schedule an additional verified cleaning cycle without a multiday exception process. As a result, one line often sits unused while another accumulates a twelve-hour queue. During allergen investigations, production planners must either cancel unaffected orders or wait for the designated line to reopen, even when inspected equipment elsewhere is available. Overtime, spoilage, and missed shipments have risen for six consecutive months. Maintenance records show no corresponding loss of mechanical capacity, and staffing levels have remained stable.","scenario_title":"Idle Filling Lines Amid Growing Backlogs","vocabulary_separation_audit":"Uses factory-specific language about filling lines, sanitation cycles, allergens, queues, spoilage, and mechanical capacity; it does not repeat the supplied predicate or name the archetype."},{"alternative_route_blocks":[],"case_id":"E14A039__TRANSFER","case_type":"TRANSFER_POSITIVE","domain":"Online marketplace promotion governance","intended_route_id":"B1","omitted_condition_id":null,"remedy_leakage_audit":"The statement identifies how campaign controls contribute to losses but offers no corrective design, intervention, or recommendation.","route_evidence":[{"condition_id":"ext001_constraint_envelope_adjustment_a01","intended_status":"SATISFIED","scenario_evidence":"Campaign permissions allow individually acceptable discounts, credits, and targeting settings to combine without aggregate limits, creating harmful commercial exposure."}],"scenario_text":"An online marketplace lets category managers launch seller-funded promotions without engineering support. Each campaign form separately accepts percentage discounts, platform credits, free shipping, regional targeting, and eligibility rules. The authorization checks validate each field but do not evaluate the combined financial exposure across overlapping campaigns. During a holiday weekend, three managers independently promoted the same electronics inventory: one applied a 25 percent discount, another issued a new-customer credit, and a third enabled free international shipping. All submissions passed their assigned approval thresholds. When the offers stacked, thousands of orders were priced below fulfillment cost, with most losses concentrated in two high-freight countries. Similar overlaps have occurred four times this quarter as campaign volume has grown. Managers cannot see pending promotions owned by other categories before publication, while finance receives exposure totals only after transactions settle. No fraud or unauthorized account use was found; the losses arose from actions available to ordinary campaign owners.","scenario_title":"Stacked Promotions Turn Sales Into Losses","vocabulary_separation_audit":"Shifts to digital commerce and uses campaign forms, discounts, credits, shipping, targeting, approvals, and settlement vocabulary, substantially distinct from the manufacturing case."},{"alternative_route_blocks":[],"case_id":"E14A039__NEAR_MISS","case_type":"ONE_LITERAL_NEAR_MISS","domain":"Online marketplace promotion governance","intended_route_id":"B1","omitted_condition_id":"ext001_constraint_envelope_adjustment_a01","remedy_leakage_audit":"The statement describes evidence about the incident's source without suggesting changes to feeds, controls, software, or review procedures.","route_evidence":[{"condition_id":"ext001_constraint_envelope_adjustment_a01","intended_status":"CONTRADICTED","scenario_evidence":"The incident review establishes that campaign permissions prevented excessive combinations, allowed all legitimate launches, distributed exposure normally, and neither caused losses nor impeded response."}],"scenario_text":"An online marketplace experienced a holiday-weekend loss on discounted electronics sold into two high-freight countries. At first, investigators suspected that several seller promotions had combined improperly. The campaign ledger, however, shows that the platform rejected every attempted overlap exceeding the established per-order ceiling. Approved discounts, customer credits, and shipping offers remained within both individual and aggregate limits. All legitimate campaigns launched on schedule, no category manager was prevented from pausing an offer, and exposure was distributed normally across regions and teams. The loss instead appears only on orders whose landed-cost estimate used a currency table imported at 02:00 Saturday. For ninety-seven minutes, that table labeled minor units as major units for one currency pair, understating fulfillment expense before campaign evaluation began. Replaying the same orders with the recorded promotion settings and the prior currency table produces positive margins. Access logs also show no unauthorized activity, delayed approval, or unavailable campaign action during the incident.","scenario_title":"Promotion Controls Cleared After Pricing Incident","vocabulary_separation_audit":"Retains the transfer case's commerce setting and analytical complexity while using incident-ledger, currency-table, replay, and access-log evidence to distinguish the near miss and explicitly negate the route condition."}],"experiment_id":"eoa_inverse_innovation_exp14_applicability_retrieval40_20260813","sample_id":"E14A039","schema_version":1}