CAIN-42 CAIN Studio

Evidence library · family

Bench category

336 tested rules in the 'bench_category' family, 336 held.

Last reviewed 2026-10-01

336 of 336 held   niche Benchmarks, coverage & performance

What this family tests

Every rule the CAIN-42 test suites recorded under the family bench_category, across 8 evidence bundles. Each rule links to its own page with the recorded result and an in-browser check of the file it came from.

IDRuleBundleResult
E32-CAT-self_improvementevery self_improvement scenario is heldE32held
E32-CAT-reward_hackingevery reward_hacking scenario is heldE32held
E32-CAT-evaluator_manipulationevery evaluator_manipulation scenario is heldE32held
E32-CAT-world_model_poisoningevery world_model_poisoning scenario is heldE32held
E32-CAT-memory_poisoningevery memory_poisoning scenario is heldE32held
E32-CAT-evidence_poisoningevery evidence_poisoning scenario is heldE32held
E32-CAT-authority_driftevery authority_drift scenario is heldE32held
E32-CAT-policy_driftevery policy_drift scenario is heldE32held
E32-CAT-identity_driftevery identity_drift scenario is heldE32held
E32-CAT-model_substitutionevery model_substitution scenario is heldE32held
E32-CAT-runtime_substitutionevery runtime_substitution scenario is heldE32held
E32-CAT-agent_substitutionevery agent_substitution scenario is heldE32held
E32-CAT-delegation_attacksevery delegation_attacks scenario is heldE32held
E32-CAT-collective_attacksevery collective_attacks scenario is heldE32held
E32-CAT-supply_chainevery supply_chain scenario is heldE32held
E32-CAT-protocol_attacksevery protocol_attacks scenario is heldE32held
E32-CAT-proof_attacksevery proof_attacks scenario is heldE32held
E32-CAT-verifier_attacksevery verifier_attacks scenario is heldE32held
E32-CAT-rollback_attacksevery rollback_attacks scenario is heldE32held
E32-CAT-recovery_attacksevery recovery_attacks scenario is heldE32held
E32-CAT-recursive_attacksevery recursive_attacks scenario is heldE32held
E32-CAT-governance_loop_attacksevery governance_loop_attacks scenario is heldE32held
E33-CAT-identityevery identity scenario holdsE33held
E33-CAT-authorityevery authority scenario holdsE33held
E33-CAT-delegationevery delegation scenario holdsE33held
E33-CAT-capabilityevery capability scenario holdsE33held
E33-CAT-policyevery policy scenario holdsE33held
E33-CAT-memoryevery memory scenario holdsE33held
E33-CAT-evidenceevery evidence scenario holdsE33held
E33-CAT-proofevery proof scenario holdsE33held
E33-CAT-verifierevery verifier scenario holdsE33held
E33-CAT-communicationevery communication scenario holdsE33held
E33-CAT-collectiveevery collective scenario holdsE33held
E33-CAT-economicevery economic scenario holdsE33held
E33-CAT-transactionevery transaction scenario holdsE33held
E33-CAT-supply_chainevery supply_chain scenario holdsE33held
E33-CAT-runtimeevery runtime scenario holdsE33held
E33-CAT-modelevery model scenario holdsE33held
E33-CAT-world_modelevery world_model scenario holdsE33held
E33-CAT-physical_actionevery physical_action scenario holdsE33held
E33-CAT-computer_useevery computer_use scenario holdsE33held
E33-CAT-self_improvementevery self_improvement scenario holdsE33held
E33-CAT-governance_loopevery governance_loop scenario holdsE33held
E33-CAT-recursiveevery recursive scenario holdsE33held
E33-CAT-recoveryevery recovery scenario holdsE33held
E33-CAT-partitionevery partition scenario holdsE33held
E33-CAT-byzantineevery byzantine scenario holdsE33held
E33-CAT-insiderevery insider scenario holdsE33held
E34-CAT-envelopeevery envelope scenario holdsE34held
E34-CAT-statusevery status scenario holdsE34held
E34-CAT-receipt_chainevery receipt_chain scenario holdsE34held
E34-CAT-delegation_proofevery delegation_proof scenario holdsE34held
E34-CAT-coverageevery coverage scenario holdsE34held
E34-CAT-conformanceevery conformance scenario holdsE34held
E34-CAT-certificateevery certificate scenario holdsE34held
E34-CAT-contractevery contract scenario holdsE34held
E34-CAT-disclosureevery disclosure scenario holdsE34held
E34-CAT-federationevery federation scenario holdsE34held
E34-CAT-edgeevery edge scenario holdsE34held
E34-CAT-transactionevery transaction scenario holdsE34held
E34-CAT-computer_useevery computer_use scenario holdsE34held
E34-CAT-evolutionevery evolution scenario holdsE34held
E34-CAT-supply_chainevery supply_chain scenario holdsE34held
E34-CAT-recoveryevery recovery scenario holdsE34held
E34-CAT-reputationevery reputation scenario holdsE34held
E34-CAT-governabilityevery governability scenario holdsE34held
E34-CAT-tenancyevery tenancy scenario holdsE34held
E34-CAT-handshakeevery handshake scenario holdsE34held
E34-CAT-interchangeevery interchange scenario holdsE34held
E34-CAT-immune_v3every immune_v3 scenario holdsE34held
E34-CAT-failure_to_proofevery failure_to_proof scenario holdsE34held
E34-CAT-replayevery replay scenario holdsE34held
E34-CAT-time_machineevery time_machine scenario holdsE34held
E34-CAT-dataplaneevery dataplane scenario holdsE34held
E34-CAT-truth_guardevery truth_guard scenario holdsE34held
E34-CAT-moatevery moat scenario holdsE34held
E34-CAT-lawevery law scenario holdsE34held
E34-CAT-primitiveevery primitive scenario holdsE34held
E34-CAT-enforcementevery enforcement scenario holdsE34held
E34-CAT-proof_exchangeevery proof_exchange scenario holdsE34held
E35-CAT-lawevery law scenario holdsE35held
E35-CAT-epistemicevery epistemic scenario holdsE35held
E35-CAT-observationevery observation scenario holdsE35held
E35-CAT-riskevery risk scenario holdsE35held
E35-CAT-predictionevery prediction scenario holdsE35held
E35-CAT-calibrationevery calibration scenario holdsE35held
E35-CAT-recommendationevery recommendation scenario holdsE35held
E35-CAT-policy_optimizerevery policy_optimizer scenario holdsE35held
E35-CAT-control_optimizerevery control_optimizer scenario holdsE35held
E35-CAT-learnerevery learner scenario holdsE35held
E35-CAT-world_modelevery world_model scenario holdsE35held
E35-CAT-counterfactualevery counterfactual scenario holdsE35held
E35-CAT-simulationevery simulation scenario holdsE35held
E35-CAT-evaluationevery evaluation scenario holdsE35held
E35-CAT-benchmarkevery benchmark scenario holdsE35held
E35-CAT-memoryevery memory scenario holdsE35held
E35-CAT-knowledgeevery knowledge scenario holdsE35held
E35-CAT-researchevery research scenario holdsE35held
E35-CAT-environment_queryevery environment_query scenario holdsE35held
E35-CAT-state_predictorevery state_predictor scenario holdsE35held
E35-CAT-consequence_graphevery consequence_graph scenario holdsE35held
E35-CAT-autonomy_gradientevery autonomy_gradient scenario holdsE35held
E35-CAT-budget_learnerevery budget_learner scenario holdsE35held
E35-CAT-harnessevery harness scenario holdsE35held
E35-CAT-plan_check_actevery plan_check_act scenario holdsE35held
E35-CAT-goal_integrityevery goal_integrity scenario holdsE35held
E35-CAT-intent_compilerevery intent_compiler scenario holdsE35held
E35-CAT-action_graphevery action_graph scenario holdsE35held
E35-CAT-trajectoryevery trajectory scenario holdsE35held
E35-CAT-communicationevery communication scenario holdsE35held
E35-CAT-collusionevery collusion scenario holdsE35held
E35-CAT-economyevery economy scenario holdsE35held
E35-CAT-laborevery labor scenario holdsE35held
E35-CAT-marketplaceevery marketplace scenario holdsE35held
E35-CAT-compilerevery compiler scenario holdsE35held
E35-CAT-obligationevery obligation scenario holdsE35held
E35-CAT-policy_conflictevery policy_conflict scenario holdsE35held
E35-CAT-failure_predictorevery failure_predictor scenario holdsE35held
E35-CAT-failure_to_productevery failure_to_product scenario holdsE35held
E35-CAT-adversarial_generationevery adversarial_generation scenario holdsE35held
E35-CAT-red_teamevery red_team scenario holdsE35held
E35-CAT-blue_teamevery blue_team scenario holdsE35held
E35-CAT-tournamentevery tournament scenario holdsE35held
E35-CAT-regretevery regret scenario holdsE35held
E35-CAT-calibration_engineevery calibration_engine scenario holdsE35held
E35-CAT-deploymentevery deployment scenario holdsE35held
E35-CAT-profileevery profile scenario holdsE35held
E35-CAT-computer_useevery computer_use scenario holdsE35held
E35-CAT-revisionevery revision scenario holdsE35held
E35-CAT-r2eevery r2e scenario holdsE35held
E35-CAT-radarevery radar scenario holdsE35held
E35-CAT-ipevery ip scenario holdsE35held
E35-CAT-moatevery moat scenario holdsE35held
E35-CAT-swarmevery swarm scenario holdsE35held
E35-CAT-labevery lab scenario holdsE35held
E35-CAT-benchmark_v2every benchmark_v2 scenario holdsE35held
E35-CAT-integrityevery integrity scenario holdsE35held
E35-CAT-root_of_trustevery root_of_trust scenario holdsE35held
E35-CAT-degradationevery degradation scenario holdsE35held
E35-CAT-latencyevery latency scenario holdsE35held
E35-CAT-economicsevery economics scenario holdsE35held
E35-CAT-self_healingevery self_healing scenario holdsE35held
E35-CAT-self_governanceevery self_governance scenario holdsE35held
E35-CAT-evolutionevery evolution scenario holdsE35held
E35-CAT-canaryevery canary scenario holdsE35held
E35-CAT-rollbackevery rollback scenario holdsE35held
E35-CAT-loopevery loop scenario holdsE35held
E35-CAT-fabricevery fabric scenario holdsE35held
E36-CAT-lawevery law scenario holdsE36held
E36-CAT-identityevery identity scenario holdsE36held
E36-CAT-passportevery passport scenario holdsE36held
E36-CAT-discoveryevery discovery scenario holdsE36held
E36-CAT-matchevery match scenario holdsE36held
E36-CAT-marketplaceevery marketplace scenario holdsE36held
E36-CAT-procurementevery procurement scenario holdsE36held
E36-CAT-negotiationevery negotiation scenario holdsE36held
E36-CAT-contractevery contract scenario holdsE36held
E36-CAT-contract_compileevery contract_compile scenario holdsE36held
E36-CAT-slaevery sla scenario holdsE36held
E36-CAT-performanceevery performance scenario holdsE36held
E36-CAT-reputationevery reputation scenario holdsE36held
E36-CAT-escrowevery escrow scenario holdsE36held
E36-CAT-paymentevery payment scenario holdsE36held
E36-CAT-receiptevery receipt scenario holdsE36held
E36-CAT-disputeevery dispute scenario holdsE36held
E36-CAT-arbitrationevery arbitration scenario holdsE36held
E36-CAT-risk_poolevery risk_pool scenario holdsE36held
E36-CAT-warrantyevery warranty scenario holdsE36held
E36-CAT-proof_marketevery proof_market scenario holdsE36held
E36-CAT-bidevery bid scenario holdsE36held
E36-CAT-auctionevery auction scenario holdsE36held
E36-CAT-compositionevery composition scenario holdsE36held
E36-CAT-supply_chainevery supply_chain scenario holdsE36held
E36-CAT-subcontractevery subcontract scenario holdsE36held
E36-CAT-service_chain_proofevery service_chain_proof scenario holdsE36held
E36-CAT-quality_graphevery quality_graph scenario holdsE36held
E36-CAT-service_twinevery service_twin scenario holdsE36held
E36-CAT-procurement_simevery procurement_sim scenario holdsE36held
E36-CAT-econ_twinevery econ_twin scenario holdsE36held
E36-CAT-econ_attackevery econ_attack scenario holdsE36held
E36-CAT-firebreakevery firebreak scenario holdsE36held
E36-CAT-org_marketevery org_market scenario holdsE36held
E36-CAT-collectiveevery collective scenario holdsE36held
E36-CAT-labor_contractevery labor_contract scenario holdsE36held
E36-CAT-a2aevery a2a scenario holdsE36held
E36-CAT-directoryevery directory scenario holdsE36held
E36-CAT-trust_negotiationevery trust_negotiation scenario holdsE36held
E36-CAT-msdpevery msdp scenario holdsE36held
E36-CAT-descriptorevery descriptor scenario holdsE36held
E36-CAT-service_certevery service_cert scenario holdsE36held
E36-CAT-telemetryevery telemetry scenario holdsE36held
E36-CAT-econ_graphevery econ_graph scenario holdsE36held
E36-CAT-econ_knowledgeevery econ_knowledge scenario holdsE36held
E36-CAT-corpusevery corpus scenario holdsE36held
E36-CAT-threat_intelevery threat_intel scenario holdsE36held
E36-CAT-service_searchevery service_search scenario holdsE36held
E36-CAT-procurement_agentevery procurement_agent scenario holdsE36held
E36-CAT-substitutionevery substitution scenario holdsE36held
E36-CAT-continuityevery continuity scenario holdsE36held
E36-CAT-routerevery router scenario holdsE36held
E36-CAT-econ_policyevery econ_policy scenario holdsE36held
E36-CAT-gov_budgetevery gov_budget scenario holdsE36held
E36-CAT-safe_modeevery safe_mode scenario holdsE36held
E36-CAT-moatevery moat scenario holdsE36held
E36-CAT-agency_graphevery agency_graph scenario holdsE36held
E36-CAT-competitionevery competition scenario holdsE36held
E36-CAT-service_evolutionevery service_evolution scenario holdsE36held
E36-CAT-revocationevery revocation scenario holdsE36held
E36-CAT-recoveryevery recovery scenario holdsE36held
E36-CAT-internalevery internal scenario holdsE36held
E36-CAT-competitive_gapevery competitive_gap scenario holdsE36held
E36-CAT-exchangeevery exchange scenario holdsE36held
E40-CAT-lawevery law scenario holdsE40held
E40-CAT-epistemicevery epistemic scenario holdsE40held
E40-CAT-world_modelevery world_model scenario holdsE40held
E40-CAT-reconcileevery reconcile scenario holdsE40held
E40-CAT-driftevery drift scenario holdsE40held
E40-CAT-causalevery causal scenario holdsE40held
E40-CAT-objectiveevery objective scenario holdsE40held
E40-CAT-riskevery risk scenario holdsE40held
E40-CAT-resourceevery resource scenario holdsE40held
E40-CAT-authorityevery authority scenario holdsE40held
E40-CAT-forecastevery forecast scenario holdsE40held
E40-CAT-calibrationevery calibration scenario holdsE40held
E40-CAT-counterfactualevery counterfactual scenario holdsE40held
E40-CAT-simulationevery simulation scenario holdsE40held
E40-CAT-autonomy_budgetevery autonomy_budget scenario holdsE40held
E40-CAT-autonomy_gradientevery autonomy_gradient scenario holdsE40held
E40-CAT-humanevery human scenario holdsE40held
E40-CAT-checkpointevery checkpoint scenario holdsE40held
E40-CAT-completionevery completion scenario holdsE40held
E40-CAT-false_completionevery false_completion scenario holdsE40held
E40-CAT-clockevery clock scenario holdsE40held
E40-CAT-degradationevery degradation scenario holdsE40held
E40-CAT-replanevery replan scenario holdsE40held
E40-CAT-time_machineevery time_machine scenario holdsE40held
E40-CAT-proofevery proof scenario holdsE40held
E40-CAT-certificateevery certificate scenario holdsE40held
E40-CAT-loopevery loop scenario holdsE40held
E40-CAT-rejectevery reject scenario holdsE40held
E40-CAT-scaleevery scale scenario holdsE40held
E40-CAT2-lawcategory law has many held scenariosE40held
E40-CAT2-epistemiccategory epistemic has many held scenariosE40held
E40-CAT2-world_modelcategory world_model has many held scenariosE40held
E40-CAT2-reconcilecategory reconcile has many held scenariosE40held
E40-CAT2-driftcategory drift has many held scenariosE40held
E40-CAT2-causalcategory causal has many held scenariosE40held
E40-CAT2-objectivecategory objective has many held scenariosE40held
E40-CAT2-riskcategory risk has many held scenariosE40held
E40-CAT2-resourcecategory resource has many held scenariosE40held
E40-CAT2-authoritycategory authority has many held scenariosE40held
E40-CAT2-forecastcategory forecast has many held scenariosE40held
E40-CAT2-calibrationcategory calibration has many held scenariosE40held
E40-CAT2-counterfactualcategory counterfactual has many held scenariosE40held
E40-CAT2-simulationcategory simulation has many held scenariosE40held
E40-CAT2-autonomy_budgetcategory autonomy_budget has many held scenariosE40held
E40-CAT2-autonomy_gradientcategory autonomy_gradient has many held scenariosE40held
E40-CAT2-humancategory human has many held scenariosE40held
E40-CAT2-checkpointcategory checkpoint has many held scenariosE40held
E40-CAT2-completioncategory completion has many held scenariosE40held
E40-CAT2-false_completioncategory false_completion has many held scenariosE40held
E40-CAT2-clockcategory clock has many held scenariosE40held
E40-CAT2-degradationcategory degradation has many held scenariosE40held
E40-CAT2-replancategory replan has many held scenariosE40held
E40-CAT2-time_machinecategory time_machine has many held scenariosE40held
E40-CAT2-proofcategory proof has many held scenariosE40held
E40-CAT2-certificatecategory certificate has many held scenariosE40held
E40-CAT2-loopcategory loop has many held scenariosE40held
E40-CAT2-rejectcategory reject has many held scenariosE40held
E40-CAT2-scalecategory scale has many held scenariosE40held
E41-CAT-lawevery law scenario holdsE41held
E41-CAT-relationshipevery relationship scenario holdsE41held
E41-CAT-discoveryevery discovery scenario holdsE41held
E41-CAT-attestevery attest scenario holdsE41held
E41-CAT-negotiationevery negotiation scenario holdsE41held
E41-CAT-handshakeevery handshake scenario holdsE41held
E41-CAT-contractevery contract scenario holdsE41held
E41-CAT-workevery work scenario holdsE41held
E41-CAT-receiptevery receipt scenario holdsE41held
E41-CAT-settlementevery settlement scenario holdsE41held
E41-CAT-disputeevery dispute scenario holdsE41held
E41-CAT-reputationevery reputation scenario holdsE41held
E41-CAT-federationevery federation scenario holdsE41held
E41-CAT-translationevery translation scenario holdsE41held
E41-CAT-revocationevery revocation scenario holdsE41held
E41-CAT-systemicevery systemic scenario holdsE41held
E41-CAT-breakerevery breaker scenario holdsE41held
E41-CAT-sybilevery sybil scenario holdsE41held
E41-CAT-time_machineevery time_machine scenario holdsE41held
E41-CAT-threatevery threat scenario holdsE41held
E41-CAT-learningevery learning scenario holdsE41held
E41-CAT-f2tevery f2t scenario holdsE41held
E41-CAT-conformanceevery conformance scenario holdsE41held
E41-CAT-rejectevery reject scenario holdsE41held
E41-CAT-scaleevery scale scenario holdsE41held
E41-CAT2-lawcategory law has held scenariosE41held
E41-CAT2-relationshipcategory relationship has held scenariosE41held
E41-CAT2-discoverycategory discovery has held scenariosE41held
E41-CAT2-attestcategory attest has held scenariosE41held
E41-CAT2-negotiationcategory negotiation has held scenariosE41held
E41-CAT2-handshakecategory handshake has held scenariosE41held
E41-CAT2-contractcategory contract has held scenariosE41held
E41-CAT2-workcategory work has held scenariosE41held
E41-CAT2-receiptcategory receipt has held scenariosE41held
E41-CAT2-settlementcategory settlement has held scenariosE41held
E41-CAT2-disputecategory dispute has held scenariosE41held
E41-CAT2-reputationcategory reputation has held scenariosE41held
E41-CAT2-federationcategory federation has held scenariosE41held
E41-CAT2-translationcategory translation has held scenariosE41held
E41-CAT2-revocationcategory revocation has held scenariosE41held
E41-CAT2-systemiccategory systemic has held scenariosE41held
E41-CAT2-breakercategory breaker has held scenariosE41held
E41-CAT2-sybilcategory sybil has held scenariosE41held
E41-CAT2-time_machinecategory time_machine has held scenariosE41held
E41-CAT2-threatcategory threat has held scenariosE41held
E41-CAT2-learningcategory learning has held scenariosE41held
E41-CAT2-f2tcategory f2t has held scenariosE41held
E41-CAT2-conformancecategory conformance has held scenariosE41held
E41-CAT2-rejectcategory reject has held scenariosE41held
E41-CAT2-scalecategory scale has held scenariosE41held
E42-CAT-lawevery law scenario holdsE42held
E42-CAT-kernelevery kernel scenario holdsE42held
E42-CAT-actionevery action scenario holdsE42held
E42-CAT-abievery abi scenario holdsE42held
E42-CAT-proofevery proof scenario holdsE42held
E42-CAT-receiptevery receipt scenario holdsE42held
E42-CAT-coverageevery coverage scenario holdsE42held
E42-CAT-registryevery registry scenario holdsE42held
E42-CAT-truthevery truth scenario holdsE42held
E42-CAT-graphevery graph scenario holdsE42held
E42-CAT-dedupevery dedup scenario holdsE42held
E42-CAT-evolutionevery evolution scenario holdsE42held
E42-CAT-time_machineevery time_machine scenario holdsE42held
E42-CAT-false_completionevery false_completion scenario holdsE42held
E42-CAT-rejectevery reject scenario holdsE42held
E42-CAT-scaleevery scale scenario holdsE42held

Other families in this niche

bench_family · coverage · property · telemetry · fabric_component · radar · swe · harness · regression · performance · corpus · end_to_end · completion · latency · f2t · false_completion · failure_to_test

Try CAIN-42 on your own agents

Create a free account and every new account starts with a 7-day trial of the full platform. Or try the sandbox first, with no account at all.

Create a free account →  ·  Try the sandbox  ·  See the whole ecosystem