{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp09_archetype_breadth150_20260804","research_id":"eoa_inverse_innovation_exp09_light_prior_art_20260804","cell_id":"computability_boundary_mapping__chemistry_materials","search_lanes":{"direct_problem_and_intervention":{"queries":["chemical reaction network reachability undecidable Petri net safety hazard species","reaction network reachability explicit unknown timeout safety model checking"],"source_ids":["SRC1","SRC2","SRC3"],"no_result_note":null},"synonyms_and_historical_terms":{"queries":["vector addition systems Petri net reachability decidability chemical reaction networks","inhibitor arcs zero checking Turing universal reachability undecidable","feed-forward autogenesis void reactions CRN reachability"],"source_ids":["SRC1","SRC2"],"no_result_note":null},"products_practices_and_standards":{"queries":["PRISM biochemical pathway finite state species upper bound simulation model checking","SMT-LIB unknown memout incomplete unsupported response standard"],"source_ids":["SRC3","SRC4"],"no_result_note":null},"component_combination":{"queries":["bounded chemical reaction network exhaustive finite state reachability","inhibitory chemical reaction networks restricted decidable subclasses","simulation maximum path length unevaluated property error model checking"],"source_ids":["SRC1","SRC2","SRC3","SRC4"],"no_result_note":null}},"sources":[{"source_id":"SRC1","title":"Reachability in Restricted Chemical Reaction Networks","publisher":"arXiv","url":"https://arxiv.org/abs/2211.12603","source_type":"PRIMARY_RESEARCH","claims_supported":["Standard discrete chemical reaction networks are equivalent to vector addition systems and Petri nets for the studied reachability problem.","General standard-CRN reachability is decidable but Ackermann-complete, while restrictions involving rule size, volume, feed-forward structure, void rules, and autogenesis produce substantially different complexity classes.","The result shows why unbounded populations or slow searches alone do not imply undecidability and why the exact reaction semantics matter."]},{"source_id":"SRC2","title":"Reachability with Restricted Reactions in Inhibitory Chemical Reaction Networks","publisher":"Schloss Dagstuhl – Leibniz-Zentrum für Informatik","url":"https://drops.dagstuhl.de/storage/00lipics/lipics-vol370-swat2026/html/LIPIcs.SWAT.2026.3/LIPIcs.SWAT.2026.3.html","source_type":"PRIMARY_RESEARCH","claims_supported":["Classical CRNs, Petri nets, and vector addition systems are equivalent infinite-state models whose ordinary reachability problem is decidable and Ackermann-complete.","Adding inhibition supplies zero-checking, can yield Turing universality, and makes general inhibitory-CRN reachability undecidable.","Restricted deletion-only and volume-preserving inhibitory systems recover decidable cases with complexity ranging from polynomial time through NP and PSPACE, directly demonstrating a semantics-dependent computability boundary."]},{"source_id":"SRC3","title":"PRISM Manual","publisher":"PRISM Model Checker","url":"https://www.prismmodelchecker.org/manual/Main/AllOnOnePage","source_type":"FIRST_PARTY_PRODUCT","claims_supported":["For biochemical-pathway model checking, PRISM requires upper bounds on species amounts to ensure a finite state space and warns that feasibility is sensitive to those bounds.","PRISM distinguishes model checking from approximate statistical model checking based on sampled paths.","For an unbounded eventuality, a finite simulated path may establish neither truth nor falsity; PRISM imposes a maximum path length and reports an error when sampled paths cannot be evaluated.","Explicit-state verification, finite-state reduction, approximate simulation, exact numerical analysis, and resource limitations are already differentiated in a mature first-party tool."]},{"source_id":"SRC4","title":"The SMT-LIB Standard: Version 2.0","publisher":"SMT-LIB Initiative","url":"https://smt-lib.org/papers/smt-lib-reference-v2.0-r10.12.21.pdf","source_type":"OFFICIAL_STANDARD","claims_supported":["A standardized solver interface distinguishes sat, unsat, unknown, unsupported, and error responses rather than forcing every attempt into a Boolean answer.","For an unknown result, the standard permits reasons including memory exhaustion and solver incompleteness.","Typed uncertainty and failure semantics are therefore established formal-methods practice, although this standard does not govern chemical safety."]}],"problem_evidence":{"status":"PARTLY_SUPPORTED","finding":"The technical problem is visible: standard CRN reachability can be decidable yet extraordinarily complex, inhibition can cross into undecidability through zero-checking, and a first-party biochemical model checker requires population bounds for finite-state verification while treating unevaluable simulated paths as errors. These sources support the danger of inferring non-reachability from resource-limited search. The screen did not find direct evidence that a deployed materials-design platform currently accepts the full proposed language and converts timeouts into SAFE, so that operational prevalence remains hypothetical.","source_ids":["SRC1","SRC2","SRC3","SRC4"]},"closest_prior_art":[{"name":"CRN reachability complexity and decidability taxonomy","source_ids":["SRC1","SRC2"],"overlap":"The papers already classify chemical-reaction-network reachability by precise rule semantics and restrictions, including a boundary between decidable standard CRNs and undecidable inhibitory CRNs and multiple easier restricted subclasses.","remaining_difference":"They do not provide the candidate's materials-safety governance artifact, deployed-language versioning, fragment-admission control, five-way operational result vocabulary, independent classification approval, or authority-routing procedure."},{"name":"PRISM bounded biochemical model checking and statistical simulation","source_ids":["SRC3"],"overlap":"PRISM already requires finite species bounds for biochemical model checking, separates verification from sampled approximation, exposes resource constraints, and reports an error when a bounded simulated path cannot decide an eventuality.","remaining_difference":"The manual does not establish a computability classification for an extensible generative reaction language, require checked reductions, or define the proposed hazard-specific output and decision-authority contract."},{"name":"SMT-LIB typed solver-result protocol","source_ids":["SRC4"],"overlap":"The standard already prevents a solver's incomplete or memory-exhausted attempt from being represented as a negative proof by providing unknown, reasons for unknown, unsupported, and error responses.","remaining_difference":"It is a general satisfiability interface, not a reaction-network boundary record, finite-fragment recognizer, chemical-model validity process, or process-safety workflow."}],"prior_art_disposition":"ADJACENT_PRIOR_ART","contrastive_claim_remaining":"For a specific deployed generative-reaction grammar, combining a versioned and independently checked computability classification with enforceable pre-analysis fragment recognition and hazard-workflow result types will prevent incomplete searches from being consumed as proof of non-reachability while preserving exact decisions for verified finite instances. The individual ideas of CRN boundary classification, bounded biochemical model checking, and explicit unknown results are established; only their claimed deployed chemistry-safety integration remains contrastive.","contrastive_claim_falsifier":"The remaining claim is falsified if broader prior art already implements this combined reaction-language boundary record, enforceable fragment gate, checked classification, typed hazard outputs, and authority routing; if the deployed grammar is shown to be an ordinary standard CRN formalism with an applicable total reachability procedure and the candidate incorrectly labels it unresolved; or if an interface audit shows incomplete, timed-out, invalid, and unsupported analyses cannot be emitted or consumed as SAFE or UNREACHABLE.","gates":{"adequate_source_search":{"status":"PASS","rationale":"The bounded search covered direct CRN reachability language, older Petri-net and vector-addition-system terminology, inhibition and zero-checking, restricted and feed-forward subclasses, a first-party biochemical model checker, and a formal solver-response standard. Exactly four retained sources from four publisher contexts were opened; two are primary research, one is first-party documentation, and one is an official standard.","source_ids":["SRC1","SRC2","SRC3","SRC4"]},"supported_problem":{"status":"PASS","rationale":"The theoretical and interface-semantics problem is supported: closely related reaction languages span decidable, extremely complex, and undecidable reachability; bounded simulation can be inconclusive; and established tools and standards preserve errors or unknowns. Direct evidence of the hypothesized materials-platform behavior is absent, warranting PARTLY_SUPPORTED rather than SUPPORTED.","source_ids":["SRC1","SRC2","SRC3","SRC4"]},"distinct_testable_claim":{"status":"PASS","rationale":"After removing established components, a specific falsifiable integration claim remains: a grammar-versioned boundary record, enforceable finite-fragment admission, reviewed classification, and typed outputs must prevent incomplete searches from appearing as non-reachability in a materials hazard workflow.","source_ids":["SRC1","SRC2","SRC3","SRC4"]},"bounded_next_test":{"status":"PASS","rationale":"A time-boxed non-production test can freeze a minimal grammar, encode synthetic bounded, feed-generating, and inhibitory cases, mechanically check fragment membership, exhaust the admitted finite state spaces, attempt a grammar-matched reduction, and audit every API-to-interface status mapping. Termination, witness validity, state-space exhaustion, reduction obligations, and improper SAFE mappings are observable.","source_ids":["SRC1","SRC2","SRC3","SRC4"]},"no_obvious_safety_or_authority_stop":{"status":"PASS","rationale":"The proposed first test uses synthetic models, changes no laboratory or plant conditions, preserves existing safety controls, and leaves operational disposition with the process-safety owner. Formal reachability does not establish chemical feasibility or real-world safety, which must remain an explicit authority boundary.","source_ids":["SRC2","SRC3"]}},"screen_survival":true,"world_novelty_boundary":"This four-source bounded public-web screen cannot establish world novelty, patentability, market size, expert acceptance, or realized value. It found close foundations for every major technical component—CRN reachability boundary taxonomies, finite-state biochemical model checking, simulation limitations, and typed unknown responses—but no retained source containing the complete materials-safety governance integration. Broader patent, product, standards, and literature searches could change the ADJACENT_PRIOR_ART disposition."}