{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp09_archetype_breadth150_20260804","research_id":"eoa_inverse_innovation_exp09_light_prior_art_20260804","cell_id":"versioning_and_quality_discrimination__computer_science","search_lanes":{"direct_problem_and_intervention":{"queries":["hosted fuzzing service crash minimization pricing tiers SLA priority support","test case minimization counterexample reduction cloud fuzzing pricing plans"],"source_ids":["SRC1","SRC2","SRC3"],"no_result_note":null},"synonyms_and_historical_terms":{"queries":["counterexample shrinking delta debugging hosted service service levels","failure-inducing input minimization fuzzing service pricing"],"source_ids":["SRC1","SRC4"],"no_result_note":null},"products_practices_and_standards":{"queries":["ClusterFuzz testcase minimization reproduction crash report","Mayhem API fuzzing pricing plans enterprise personalized support","commercial fuzzing platform pricing plans CI integration"],"source_ids":["SRC1","SRC2"],"no_result_note":null},"component_combination":{"queries":["cloud fuzzing reserved capacity priority queue CI integration human triage","fuzzing service concurrency SLA expert support testcase minimization","delay price differentiation cloud computing priority queues SLA"],"source_ids":["SRC1","SRC2","SRC3"],"no_result_note":null}},"sources":[{"source_id":"SRC1","title":"Uploading a testcase","publisher":"Google ClusterFuzz","url":"https://google.github.io/clusterfuzz/using-clusterfuzz/workflows/uploading-a-testcase/","source_type":"FIRST_PARTY_PRODUCT","claims_supported":["ClusterFuzz accepts uploaded test cases, checks whether they crash, records crash parameters and stack traces, and then performs test-case minimization and regression-range analysis.","The documented processing speed depends on bot availability, making latency and capacity operationally visible.","The workflow supports the proposal's premise that crash reproduction, reduction, and reporting consume queued hosted resources."]},{"source_id":"SRC2","title":"Introducing The Mayhem for API Free Plan","publisher":"Mayhem Security","url":"https://www.mayhem.security/blog/introducing-the-mayhem-for-api-free-plan","source_type":"FIRST_PARTY_PRODUCT","claims_supported":["Mayhem offered a free fuzzing plan capped at 50 monthly scans after feedback from solo developers and startups seeking a suitably sized plan.","Customers needing additional scans, enterprise features, or personalized support could upgrade to paid plans.","This establishes prior commercial bundling of fuzzing capacity, features, and human support while preserving a lower-cost access path."]},{"source_id":"SRC3","title":"Delay and Price Differentiation in Cloud Computing: A Service Model, Supporting Architectures, and Performance","publisher":"arXiv","url":"https://arxiv.org/abs/2007.15314","source_type":"PRIMARY_RESEARCH","claims_supported":["The paper studies multiple cloud-service SLAs differentiated by processing delay and price for customers with heterogeneous delay sensitivity.","It develops market segmentation and incentive-compatible pricing and considers capacity allocation and priority-based queue architectures.","This is direct prior art for self-selection through priced latency tiers, though not specifically for fuzzing or counterexample reduction."]},{"source_id":"SRC4","title":"Simplifying and Isolating Failure-Inducing Input","publisher":"Saarland University Software Engineering Chair","url":"https://www.st.cs.uni-saarland.de/papers/tse2002/","source_type":"PRIMARY_RESEARCH","claims_supported":["Delta debugging systematically simplifies a failure-inducing test case to a minimal test case that preserves the failure.","The reported cases include reducing browser-crashing inputs and establish older terminology for the proposal's counterexample-reduction task.","The research addresses reduction technology rather than hosted service packaging, capacity, or pricing."]}],"problem_evidence":{"status":"PARTLY_SUPPORTED","finding":"The hosted workflow and relevant differentiation pressures are visible: ClusterFuzz documents reproduction, minimization, reporting, queued tasks, and completion speed dependent on bot availability, while Mayhem documents demand for an accessible free plan alongside paid capacity, enterprise features, and personalized support. Research also establishes heterogeneous delay sensitivity in cloud workloads. The bounded search did not directly verify the proposal's stronger factual assertions that one counterexample-reduction queue currently mixes routine and release-critical jobs, customers request informal priority exceptions, or large crash corpora are deferred.","source_ids":["SRC1","SRC2","SRC3","SRC4"]},"closest_prior_art":[{"name":"Mayhem for API free and paid plan ladder","source_ids":["SRC2"],"overlap":"A fuzzing provider already combined a usable free tier with paid increases in scan capacity, enterprise features, and personalized support, closely overlapping access preservation, usage limits, feature gating, upgrades, and human assistance.","remaining_difference":"The opened source does not describe tiers for minimizing submitted crashes or differentiate them through queue-time targets, reserved reduction capacity, retention, and an invariant clean-environment reproduction floor."},{"name":"Delay- and price-differentiated cloud SLAs","source_ids":["SRC3"],"overlap":"The research directly models customers with different delay sensitivity selecting differently priced SLAs and analyzes queues, capacity allocation, incentive compatibility, and revenue.","remaining_difference":"It is a general cloud-computing model and does not combine fuzzing-specific reduction, reproducibility, confidentiality, CI handoff, retention, or bounded human triage."},{"name":"ClusterFuzz hosted test-case processing","source_ids":["SRC1","SRC4"],"overlap":"ClusterFuzz operationalizes the established delta-debugging problem as hosted crash checking, minimization, regression analysis, and reporting whose speed depends on available bots.","remaining_difference":"The opened documentation provides no common paid Batch/Release/Incident menu, customer self-selection mechanism, reserved-capacity tokens, or tier-specific latency targets."}],"prior_art_disposition":"ADJACENT_PRIOR_ART","contrastive_claim_remaining":"For hosted crash-input reduction specifically, an identity-neutral three-version menu jointly varying disclosed queue targets, reserved capacity, concurrency, retention, CI integration, and bounded human assistance can produce comprehensible, urgency-sensitive self-selection while every version preserves the same validated reproducer, reduction correctness, confidentiality, and report-integrity floor. The retained sources disclose the technical workflow, fuzzing-plan versioning, and latency-price screening separately, but not this complete combination.","contrastive_claim_falsifier":"The contrast is falsified by finding an earlier hosted crash-reduction offering with substantially the same identity-neutral latency, reserved-capacity, integration, retention, and assistance ladder plus an invariant correctness floor, or by a blinded bounded test showing that users cannot distinguish the versions, selections do not change with standardized urgency scenarios, premium capacity is infeasible, or Batch reproducibility or access deteriorates.","gates":{"adequate_source_search":{"status":"PASS","rationale":"The bounded search covered the proposal directly, delta debugging and failure-inducing-input terminology, hosted and commercial fuzzing products, and combinations involving latency pricing, queues, capacity, integration, and support. The four opened sources span three publisher organizations and include two first-party sources and two primary-research sources.","source_ids":["SRC1","SRC2","SRC3","SRC4"]},"supported_problem":{"status":"PASS","rationale":"The problem is partly supported: authoritative product sources expose capacity-dependent hosted crash processing and commercial demand for different fuzzing capacity, features, access prices, and support, while the exact asserted queue symptoms remain unverified.","source_ids":["SRC1","SRC2","SRC3","SRC4"]},"distinct_testable_claim":{"status":"PASS","rationale":"A distinct empirical claim remains around the fuzzing-specific combination, identity-neutral selection, invariant correctness floor, and scenario-sensitive tier choice; each element can be operationally measured or contradicted.","source_ids":["SRC1","SRC2","SRC3"]},"bounded_next_test":{"status":"PASS","rationale":"The proposed two-week replay of at most 60 completed jobs plus nonbinding interviews with at most 15 consenting users can test comprehension, scenario-sensitive selection, queue feasibility, Batch-floor preservation, and guardrail burden without changing live billing or priority.","source_ids":["SRC1","SRC2","SRC3"]},"no_obvious_safety_or_authority_stop":{"status":"PASS","rationale":"A service owner can conduct the reversible offline packaging test if the security and privacy owner approves use of historical crash artifacts. Keeping confidential artifacts controlled, obtaining consent, making no live queue or billing changes, and halting if essential security functionality would be paywalled avoids an obvious initial safety or authority stop.","source_ids":["SRC1","SRC2"]}},"screen_survival":true,"world_novelty_boundary":"This result establishes only that no closer match appeared in a bounded public-web screen of the retained sources. It does not establish world novelty, patentability, non-obviousness, market size, expert acceptance, willingness to pay, operational feasibility, or realized value; patents, private contracts, discontinued products, unindexed material, and unpublished operating practices may contain closer prior art."}