# Find a retained run Choose a run to inspect its original [report, provenance and contract](lab/register.json). Each record keeps its own claim results, roles and limits. A passed claim does not establish that every claim passed or that someone independently operated the workload. Search by a project, claim, role or limit. The result filter shows records that contain the selected per-claim token. It does not give a run an overall grade. Without JavaScript, every record remains below and in the [Markdown page](runs.md). Use the [Open Evidence Lab](lab.md) for the submission protocol and review process. Use [Experiments](experiments.md) for procedures and [the claim ledger](claims.md#retained-run-navigation) for sources.
## E6-observer-admission Claim scope: verifier-behavior-on-author-fixture Review state: retained-measured-record Open [Report](experiments/observer-admission/recorded.json), [Provenance](experiments/observer-admission/provenance.json) and [Contract](https://github.com/probityai/agent-evidence-observer/blob/8562c25fb7ec97596ea9d0297c495340298914b6/docs/CONSUMER-ADMISSION.md). [Recorded procedure](experiments.md#observer-consumer-admission) ### Per-claim results | Claim | Recorded result | |-----------------------------|-----------------| | consumer-authority-binding | pass | | durable-consumer-admission | pass | | persisted-replay-refusal | pass | | signed-claim-tamper-refusal | pass | | independent-custody | not-exercised | | external-host-ci-adoption | not-exercised | | unmediated-effect-capture | not-exercised | ### Roles | Role | Declaration | |----------------------|-----------------------| | atlasReplay | author-produced | | evidenceVantage | artifact | | fixtureOperator | same-operator-fixture | | independentOperation | not-established | | witnessScope | PEER | ### Limits - One declared broker file write and one local consumer admission - No live-agent or production-isolation measurement - No independently operated witness or consumer - Consumer-state rollback can erase replay protection - No exactly-once downstream-effect claim answerExposure: Implementation, demo and retained expected outcomes were available to the atlas runner; not an answer-blind replication comparisonOrder: Retrospective retained-fixture comparison; not a result frozen before expected-answer comparison
## jep-core07-2026-10-02 Claim scope: verifier-behavior-on-public-producer-fixtures Review state: submitted-awaiting-producer-response Open [Report](experiments/jep-core07-2026-10-02/report.json), [Provenance](experiments/jep-core07-2026-10-02/provenance.json) and [Contract](https://github.com/probityai/agent-evidence-vectors/blob/4f344209c8ea65f60bab0f51747ae4f80346181d/interop/jep-core07-reader/PROTOCOL.md). ### Per-claim results | Claim | Recorded result | |-------------------------------------------------|-----------------| | validation-assertions-match-pinned-expectations | pass | | producer-assertions-match-pinned-expectations | pass | | acceptance-assertions-match-pinned-expectations | pass | | external-host-ci-adoption | not-exercised | | independent-effect-custody | not-exercised | ### Roles | Role | Declaration | |--------------------------|------------------------------| | readerAuthor | Probity | | runner | Probity-owned GitHub Actions | | producer | JEP | | independentEffectCustody | not-established | ### Limits - Pinned manifest coverage only, not complete JEP Core coverage. - Acceptance effects are synthetic local SQLite rows. - No power-loss or distributed failover experiment. answerExposure: Published expected outcomes were available before implementation; not answer-blind. comparisonOrder: Actual per-case outputs retained before comparison in memory; raw output and comparison published together, not separate revisions.
## remora-e7-2026-10-02 Claim scope: verifier-behavior-on-public-producer-fixtures Review state: producer-review-open Open [Report](experiments/remora-e7-2026-10-02/report.json), [Provenance](experiments/remora-e7-2026-10-02/provenance.json) and [Contract](https://github.com/darklordVirtual/REMORA-research/tree/e4fe474f488c3047b346abb01cbfd77ab447ad69/artifacts/interop/runtime-surface-e7-v0.1). ### Per-claim results | Claim | Recorded result | |----|----| | valid_reference_surface-matches-native-fixture-outcome | pass | | undeclared_tool_invocation_surface-matches-native-fixture-outcome | pass | | runtime_identity_changed-matches-native-fixture-outcome | pass | | observation_coverage_incomplete-matches-native-fixture-outcome | pass | | alternative_effect_path_observed-matches-native-fixture-outcome | pass | | global-runtime-capability-surface-completeness | not-exercised | | external-host-ci-adoption | not-exercised | | independent-effect-custody | not-exercised | ### Roles | Role | Declaration | |--------------------------|------------------------------| | readerAuthor | Probity | | runner | Probity-owned GitHub Actions | | producer | REMORA | | independentEffectCustody | not-established | ### Limits - Native INDEPENDENT means the external operator and separate implementation satisfy REMORA package criteria; it establishes no independent effect custody. - Global runtime_capability_surface_completeness remains NOT_ESTABLISHED. - Producer review is open; it does not establish producer acceptance, final verification or CI adoption. answerExposure: Published expected outcomes were available before implementation; not answer-blind. REMORA reference-verifier text was also exposed during lint diagnostics after implementation. comparisonOrder: Actual per-case outputs retained before comparison in memory; raw output and comparison published together, not separate revisions.
## langgraph-durable-2026-10-02 Claim scope: author-operated-native-framework-control Review state: author-retained-measured-record Open [Report](experiments/langgraph-durable-2026-10-02/report.json), [Provenance](experiments/langgraph-durable-2026-10-02/provenance.json) and [Contract](https://github.com/probityai/agent-evidence-observer/tree/90cd62c0ee926b456aa9aada3537f36ee5f17cd3/interop/langgraph-ticket-2026-10-02). ### Per-claim results | Claim | Recorded result | |---------------------------------------------|-----------------| | bounded-six-case-population | pass | | reader-reconstructed-native-record | pass | | crash-after-effect-recovers-at-revision-one | pass | | pending-intent-no-automatic-replay | pass | | missing-checkpoint-refusal | pass | | wrong-thread-refusal | pass | | external-host-recurring-adoption | not-exercised | | independent-effect-custody | not-exercised | ### Roles | Role | Declaration | |--------------------------|------------------------| | producer | Probity-owned workflow | | readerAuthor | Probity | | atlasOperator | Probity | | retentionHolder | Probity | | independentOperation | not-established | | independentEffectCustody | not-established | | outsideRecurringUse | not-established | ### Limits - Graph-worker hard exits/reopened SQLite checkpoints only; target process remains alive. - No target-process restart, host power loss, issuer authentication, general exactly-once guarantee or independent custody. answerExposure: Author-written tasks/controls and expected outcomes exposed; not blind. comparisonOrder: Native output and comparison published together; not separate raw-result-first revisions.
## pydantic-failure-2026-10-02 Claim scope: author-operated-native-framework-control Review state: author-retained-measured-record Open [Report](experiments/pydantic-failure-2026-10-02/report.json), [Provenance](experiments/pydantic-failure-2026-10-02/provenance.json) and [Contract](https://github.com/probityai/agent-evidence-observer/tree/afda549c3f20f8e7b031275c06558a50bf20726e/interop/pydantic-ai-native-2026-10-02). ### Per-claim results | Claim | Recorded result | |---------------------------------------------|-----------------| | bounded-seven-case-population | pass | | reader-reconstructed-native-record | pass | | retry-exhaustion-remains-error-no-effect | pass | | committed-effect-error-retains-revision-one | pass | | external-host-recurring-adoption | not-exercised | | independent-effect-custody | not-exercised | ### Roles | Role | Declaration | |--------------------------|------------------------| | producer | Probity-owned workflow | | readerAuthor | Probity | | atlasOperator | Probity | | retentionHolder | Probity | | independentOperation | not-established | | independentEffectCustody | not-established | | outsideRecurringUse | not-established | ### Limits - Real Pydantic framework with scripted FunctionModel; no model-quality or real-provider inference measurement. - Selected loopback ticket effects; no caller authentication, outside adoption or independent custody. answerExposure: Author-written tasks/controls and expected outcomes exposed; not blind. comparisonOrder: Native output and comparison published together; not separate raw-result-first revisions.
## model-operational-2026-10-02 Claim scope: author-operated-actual-weight-microtask-result Review state: author-retained-measured-record Open [Report](experiments/model-operational-2026-10-02/report.json), [Provenance](experiments/model-operational-2026-10-02/provenance.json) and [Contract](https://github.com/probityai/agent-evidence-observer/tree/ab84161d09bb7bd696bba57717a418a95dd1d976/interop/local-model-tasks-2026-10-02). ### Per-claim results | Claim | Recorded result | |------------------------------------------------|-----------------| | complete-declared-population | pass | | evidence-complete | pass | | declared-run-budget | pass | | task-correctness-short24-structured-extraction | fail | | task-correctness-short24-arithmetic | fail | | task-correctness-short24-policy-decision | fail | | task-correctness-short24-grounded-abstention | fail | | task-correctness-long96-structured-extraction | fail | | task-correctness-long96-arithmetic | fail | | task-correctness-long96-policy-decision | fail | | task-correctness-long96-grounded-abstention | fail | | external-host-recurring-adoption | not-exercised | | independent-effect-custody | not-exercised | ### Roles | Role | Declaration | |--------------------------|------------------------| | producer | Probity-owned workflow | | readerAuthor | Probity | | atlasOperator | Probity | | retentionHolder | Probity | | independentOperation | not-established | | independentEffectCustody | not-established | | outsideRecurringUse | not-established | ### Limits - Original 274795286-byte ZIP retained locally and available through expiring GitHub workflow artifact, not committed to Git or represented by the capsule. - Capsule retains exact selected JSON/text members and excludes model weights/binaries; it cannot replay the full source/binary integrity check or rerun inference alone. - Fixed configuration order confounds cache/warmup; author microtasks are not a benchmark estimate. - CPU is whole-process delta; RSS process-lifetime peak; no task memory isolation, monetary cost or independent effects. answerExposure: Author-written tasks/controls and expected outcomes exposed; not blind. comparisonOrder: Native output and comparison published together; not separate raw-result-first revisions.
## authority-recovery-2026-10-02 Claim scope: author-operated-native-recovery-authority-control Review state: author-retained-measured-record Open [Report](experiments/authority-recovery-2026-10-02/report.json), [Provenance](experiments/authority-recovery-2026-10-02/provenance.json) and [Contract](https://github.com/probityai/agent-evidence-observer/tree/b02988a771bea3652660bec934c3c599c3ec4d27/interop/langgraph-recovery-authority-2026-10-02). ### Per-claim results | Claim | Recorded result | |----------------------------------------------|-----------------| | complete-selected-population | pass | | exact-reader-reconstruction | pass | | valid-before-recovery-and-effect-separate | pass | | valid-after-recovery-and-effect-separate | pass | | revoked-before-recovery-and-effect-separate | pass | | revoked-after-recovery-and-effect-separate | pass | | expired-before-recovery-and-effect-separate | pass | | expired-after-recovery-and-effect-separate | pass | | rollback-before-recovery-and-effect-separate | pass | | rollback-after-recovery-and-effect-separate | pass | | external-host-recurring-adoption | not-exercised | | independent-effect-custody | not-exercised | ### Roles | Role | Declaration | |--------------------------|------------------------| | producer | Probity-owned workflow | | readerAuthor | Probity | | atlasOperator | Probity | | retentionHolder | Probity | | independentOperation | not-established | | independentEffectCustody | not-established | | outsideRecurringUse | not-established | ### Limits - Host-selected deterministic integer recovery clock; signed target grant remains separate. - Graph workers hard-exit/reopen; target process stays alive; no host power loss, distributed exactly-once or independent custody. answerExposure: Author-written tasks/controls and expected outcomes exposed; not blind. comparisonOrder: Native output and comparison emitted and published together; public Atlas retention before indexing does not establish a blind or raw-result-first comparison.
## target-process-recovery-2026-10-02 Claim scope: author-operated-native-target-process-control Review state: author-retained-measured-record Open [Report](experiments/target-process-recovery-2026-10-02/report.json), [Provenance](experiments/target-process-recovery-2026-10-02/provenance.json) and [Contract](https://github.com/probityai/agent-evidence-observer/tree/37d1dd42239fbe2ef55f625855e385d7337572c1/interop/target-recovery-2026-10-02). ### Per-claim results | Claim | Recorded result | |--------------------------------------|-----------------| | complete-selected-population | pass | | exact-reader-reconstruction | pass | | retained-completedEffects | pass | | retained-incompleteRefusals | pass | | retained-startupRefusals | pass | | retained-targetProcesses | pass | | restart-ready-retained-state | pass | | crash-after-intent-retained-state | pass | | crash-inside-effect-retained-state | pass | | crash-after-effect-retained-state | pass | | concurrent-intent-retained-state | pass | | missing-store-retained-state | pass | | changed-configuration-retained-state | pass | | external-host-recurring-adoption | not-exercised | | independent-effect-custody | not-exercised | ### Roles | Role | Declaration | |--------------------------|------------------------| | producer | Probity-owned workflow | | readerAuthor | Probity | | atlasOperator | Probity | | retentionHolder | Probity | | independentOperation | not-established | | independentEffectCustody | not-established | | outsideRecurringUse | not-established | ### Limits - Selected target-process hard exits/reopen and local concurrency only; process IDs are unsigned runner testimony. - Native SQLite/key/clock/HTTP remain same-operator; no power loss, production containment, general exactly-once or outside adoption. answerExposure: Author-written tasks/controls and expected outcomes exposed; not blind. comparisonOrder: Native output and comparison emitted and published together; public Atlas retention before indexing does not establish a blind or raw-result-first comparison.
## model-paired-cpu-2026-10-02 Claim scope: author-operated-actual-weight-paired-microtask-result Review state: author-retained-measured-record Open [Report](experiments/model-paired-cpu-2026-10-02/report.json), [Provenance](experiments/model-paired-cpu-2026-10-02/provenance.json) and [Contract](https://github.com/probityai/agent-evidence-observer/tree/b7db63c319d53b0cb27638a229f966d477a2bbd5/interop/local-model-comparison-2026-10-02). ### Per-claim results | Claim | Recorded result | |----|----| | complete-selected-population | pass | | exact-reader-reconstruction | pass | | within-declared-native-run-budget | pass | | smol135-q4-short24-structured-extraction-all-six-strict-targets | fail | | smol135-q4-short24-arithmetic-all-six-strict-targets | fail | | smol135-q4-short24-policy-decision-all-six-strict-targets | fail | | smol135-q4-short24-grounded-abstention-all-six-strict-targets | fail | | smol135-q4-long96-structured-extraction-all-six-strict-targets | fail | | smol135-q4-long96-arithmetic-all-six-strict-targets | fail | | smol135-q4-long96-policy-decision-all-six-strict-targets | fail | | smol135-q4-long96-grounded-abstention-all-six-strict-targets | fail | | smol360-q4-short24-structured-extraction-all-six-strict-targets | fail | | smol360-q4-short24-arithmetic-all-six-strict-targets | fail | | smol360-q4-short24-policy-decision-all-six-strict-targets | fail | | smol360-q4-short24-grounded-abstention-all-six-strict-targets | fail | | smol360-q4-long96-structured-extraction-all-six-strict-targets | fail | | smol360-q4-long96-arithmetic-all-six-strict-targets | fail | | smol360-q4-long96-policy-decision-all-six-strict-targets | fail | | smol360-q4-long96-grounded-abstention-all-six-strict-targets | fail | | external-host-recurring-adoption | not-exercised | | independent-effect-custody | not-exercised | ### Roles | Role | Declaration | |--------------------------|------------------------| | producer | Probity-owned workflow | | readerAuthor | Probity | | atlasOperator | Probity | | retentionHolder | Probity | | independentOperation | not-established | | independentEffectCustody | not-established | | outsideRecurringUse | not-established | ### Limits - Producer-selected compact original retains native source/library/call bytes; full preparation weights/source/dependency wheels remain privately archived and provider-expiring. - Two separately declared preparation attempts, initial failure before inference and corrected preparation; cumulative transfer is not one 512MiB budget. - Author tasks/answers exposed; blocked interleaving and cache reset do not make a representative benchmark or independent operation. - Whole-process CPU and shared lifetime RSS are not per-task allocations; no live decision effects. answerExposure: Author-written tasks/controls and expected outcomes exposed; not blind. comparisonOrder: Native output and comparison emitted and published together; public Atlas retention before indexing does not establish a blind or raw-result-first comparison.
## openai-agents-ticket-2026-10-02 Claim scope: author-operated-scripted-native-sdk-control Review state: author-retained-measured-record Open [Report](experiments/openai-agents-ticket-2026-10-02/report.json), [Provenance](experiments/openai-agents-ticket-2026-10-02/provenance.json) and [Contract](https://github.com/probityai/agent-evidence-observer/tree/7a4b26058e54f0e6bbbaeab400dcd93c47607389/interop/openai-agents-ticket-2026-10-02). ### Per-claim results | Claim | Recorded result | |------------------------------------------------------|-----------------| | complete-selected-population | pass | | exact-reader-reconstruction | pass | | typed-quality-not-evaluated | pass | | scripted-sdk-call-and-resource-accounting | pass | | permit-task-effect-and-resource-axes | pass | | deny-task-effect-and-resource-axes | pass | | changed-arguments-task-effect-and-resource-axes | pass | | producer-error-task-effect-and-resource-axes | pass | | committed-effect-error-task-effect-and-resource-axes | pass | | turns-exhausted-task-effect-and-resource-axes | pass | | external-host-recurring-adoption | not-exercised | | independent-effect-custody | not-exercised | ### Roles | Role | Declaration | |--------------------------|------------------------| | producer | Probity-owned workflow | | readerAuthor | Probity | | atlasOperator | Probity | | retentionHolder | Probity | | independentOperation | not-established | | independentEffectCustody | not-established | | outsideRecurringUse | not-established | ### Limits - Actual OpenAI Agents SDK Runner/function_tool/TraceProcessor with scripted Model; nine model calls do not measure remote inference or model quality. - Producer implementation, PR synthetic execution, final reader and separately pinned protected target revisions remain distinct. - Same-operator synthetic HTTP ticket effect/key/clock; no independent custody, target restart or outside recurring adoption. answerExposure: Author-written tasks/controls and expected outcomes exposed; not blind. comparisonOrder: Native output and comparison emitted and published together; public Atlas retention before indexing does not establish a blind or raw-result-first comparison.
## execsurface-state-2026-10-02 Claim scope: author-operated-native-metadata-and-endpoint-state-join Review state: author-retained-measured-record Open [Report](experiments/execsurface-state-2026-10-02/report.json), [Provenance](experiments/execsurface-state-2026-10-02/provenance.json) and [Contract](https://github.com/probityai/agent-evidence-observer/tree/88f475da4c4626b52ab78900e03ba9f73aa35640/interop/execsurface-state-2026-10-02). ### Per-claim results | Claim | Recorded result | |---------------------------------------------------|-----------------| | selected-byte-predicate-valid | pass | | selected-evidence-valid | pass | | declared-carried-write-count | pass | | separate-endpoint-after-byte-count | pass | | native-invocation-complete | pass | | import-origin-remains-disclosed | pass | | voluntary-tier-remains-disclosed | pass | | complete-observation-scope | fail | | unknown-typed-collection-health-remains-disclosed | pass | | native-write-count-not-invented | pass | | producer-calibration-pass | pass | | producer-comparison-outside-claimed-interval | pass | | independent-effect-key-store-clock-custody | not-exercised | | external-host-recurring-adoption | not-exercised | | typed-collection-health-envelope | not-exercised | | absence-of-intermediate-writes | not-exercised | ### Roles | Role | Declaration | |----|----| | producer | Probity-owned workflow using pinned ExecSurface binary | | companionAuthor | Probity | | atlasOperator | Probity | | retentionHolder | Probity | | independentOperation | not-established | | independentEffectCustody | not-established | | outsideRecurringUse | not-established | ### Limits - Voluntary/peer/log-import/software-only predicate with incomplete scope and unknown typed collection health. - Native schema-v2 metadata does not carry written bytes or write count; nine-byte content comes from separate snapshots and declared workload. - Same operator controls generated fixture, key, store, clock and captures; no independent custody, principal authentication or outside adoption. - Selected Linux x86_64 ptrace metadata only; no mmap/io_uring attribution or absence-of-intermediate-write claim. - Native learn/check producer PASS is a separate outside-interval same-byte calibration, not authority, collection health or task quality. - Release source/binary association is pinned but Rust build reproducibility is not established. answerExposure: Author-written deterministic overwrite, companion and expected outcomes exposed; not blind. comparisonOrder: Native companion output and verifier comparison emitted together; public raw retention before indexing does not establish a blind or native raw-result-first comparison.
## model-format-cpu-2026-10-02 Claim scope: author-operated-actual-weight-paired-format-and-correctness-result Review state: author-retained-measured-record Open [Report](experiments/model-format-cpu-2026-10-02/report.json), [Provenance](experiments/model-format-cpu-2026-10-02/provenance.json) and [Contract](https://github.com/probityai/agent-evidence-observer/tree/75797cbb4fafc0669849357dd2275825b469c1f9/interop/local-model-format-control-2026-10-02). ### Per-claim results | Claim | Recorded result | |----|----| | complete-selected-population | pass | | exact-reader-reconstruction | pass | | within-declared-native-run-budget | pass | | smol135-q4-short24-unconstrained-structured-extraction-all-six-strict-targets | fail | | smol135-q4-short24-unconstrained-arithmetic-all-six-strict-targets | fail | | smol135-q4-short24-unconstrained-policy-decision-all-six-strict-targets | fail | | smol135-q4-short24-unconstrained-grounded-abstention-all-six-strict-targets | fail | | smol135-q4-short24-schema-structured-extraction-all-six-strict-targets | fail | | smol135-q4-short24-schema-arithmetic-all-six-strict-targets | fail | | smol135-q4-short24-schema-policy-decision-all-six-strict-targets | fail | | smol135-q4-short24-schema-grounded-abstention-all-six-strict-targets | fail | | smol135-q4-long96-unconstrained-structured-extraction-all-six-strict-targets | fail | | smol135-q4-long96-unconstrained-arithmetic-all-six-strict-targets | fail | | smol135-q4-long96-unconstrained-policy-decision-all-six-strict-targets | fail | | smol135-q4-long96-unconstrained-grounded-abstention-all-six-strict-targets | fail | | smol135-q4-long96-schema-structured-extraction-all-six-strict-targets | fail | | smol135-q4-long96-schema-arithmetic-all-six-strict-targets | fail | | smol135-q4-long96-schema-policy-decision-all-six-strict-targets | fail | | smol135-q4-long96-schema-grounded-abstention-all-six-strict-targets | fail | | smol360-q4-short24-unconstrained-structured-extraction-all-six-strict-targets | fail | | smol360-q4-short24-unconstrained-arithmetic-all-six-strict-targets | fail | | smol360-q4-short24-unconstrained-policy-decision-all-six-strict-targets | fail | | smol360-q4-short24-unconstrained-grounded-abstention-all-six-strict-targets | fail | | smol360-q4-short24-schema-structured-extraction-all-six-strict-targets | pass | | smol360-q4-short24-schema-arithmetic-all-six-strict-targets | fail | | smol360-q4-short24-schema-policy-decision-all-six-strict-targets | fail | | smol360-q4-short24-schema-grounded-abstention-all-six-strict-targets | fail | | smol360-q4-long96-unconstrained-structured-extraction-all-six-strict-targets | fail | | smol360-q4-long96-unconstrained-arithmetic-all-six-strict-targets | fail | | smol360-q4-long96-unconstrained-policy-decision-all-six-strict-targets | fail | | smol360-q4-long96-unconstrained-grounded-abstention-all-six-strict-targets | fail | | smol360-q4-long96-schema-structured-extraction-all-six-strict-targets | pass | | smol360-q4-long96-schema-arithmetic-all-six-strict-targets | fail | | smol360-q4-long96-schema-policy-decision-all-six-strict-targets | fail | | smol360-q4-long96-schema-grounded-abstention-all-six-strict-targets | fail | | independent-effect-key-store-clock-custody | not-exercised | | external-host-recurring-adoption | not-exercised | | protected-decision-effects | not-exercised | ### Roles | Role | Declaration | |--------------------------|------------------------| | producer | Probity-owned workflow | | readerAuthor | Probity | | atlasOperator | Probity | | retentionHolder | Probity | | independentOperation | not-established | | independentEffectCustody | not-established | | outsideRecurringUse | not-established | ### Limits - Format/schema correctness separate from typed-exact target correctness; each model/cap/decoder/family row retained independently. - Protocol/compiler/grammars frozen before inference; selected integer-length, key-order and spacing restrictions limit generalization. - Exposed author tasks and same-operator native execution; no representative benchmark, live protected decision effects or independent custody. - Four separately declared preparation envelopes; cumulative transfer exceeds one512MiB envelope. - Original compact native provider archive retains sources/grammars/outputs and resources; full preparation weights/source/wheels remain privately archived. - CPU is whole-process delta and RSS shared lifetime peak; not per-model/task allocation. answerExposure: Author-written tasks/controls and expected outcomes exposed; not blind. comparisonOrder: Native output and comparison emitted and published together; public Atlas retention before indexing does not establish a blind or raw-result-first comparison.
## google-adk-ticket-2026-10-02 Claim scope: author-operated-native-google-adk-scripted-reference-result Review state: author-retained-measured-record Open [Report](experiments/google-adk-ticket-2026-10-02/report.json), [Provenance](experiments/google-adk-ticket-2026-10-02/provenance.json) and [Contract](https://github.com/probityai/agent-evidence-observer/tree/a6cb0b0eb767e01ad2d0813c9483c39af08163cd/interop/adk-ticket-2026-10-02). ### Per-claim results | Claim | Recorded result | |--------------------------------------------|-----------------| | bounded-reference-evidence-verified | pass | | declared-reference-population | pass | | scripted-SDK-resource-counts | pass | | model-quality-not-measured | pass | | permit-task-completion | pass | | permit-native-revision-one | pass | | deny-task-completion | pass | | deny-native-revision-one | fail | | changed-arguments-task-completion | pass | | changed-arguments-native-revision-one | fail | | unhandled-before-task-completion | fail | | unhandled-before-native-revision-one | fail | | unhandled-after-task-completion | fail | | unhandled-after-native-revision-one | pass | | handled-first-task-completion | pass | | handled-first-native-revision-one | pass | | handled-last-task-completion | pass | | handled-last-native-revision-one | pass | | exhausted-first-task-completion | fail | | exhausted-first-native-revision-one | fail | | exhausted-last-task-completion | fail | | exhausted-last-native-revision-one | fail | | returned-error-first-task-completion | pass | | returned-error-first-native-revision-one | pass | | returned-error-last-task-completion | pass | | returned-error-last-native-revision-one | pass | | incomplete-close-task-completion | fail | | incomplete-close-native-revision-one | pass | | independent-effect-key-store-clock-custody | not-exercised | | external-host-recurring-adoption | not-exercised | ### Roles | Role | Declaration | |--------------------------|------------------------| | producer | Probity-owned workflow | | readerAuthor | Probity | | atlasOperator | Probity | | retentionHolder | Probity | | independentOperation | not-established | | independentEffectCustody | not-established | | outsideRecurringUse | not-established | ### Limits - Real Google ADK nativeRunner/plugin/retry paths with scriptedBaseLlm; no remote inference or model-quality evaluation. - Typed MCPisError throughFunctionTool is not actual MCP transport. - Task complete/error/incomplete separate from signednativeRevision; after-effect error and early close retain committed effects. - Author-operated target/store/key/clock/capture and exposed expected outcomes; no independent custody/outside adoption. - Original synthetic execution source, producer implementation, reviewed reader and protected target baseline separately pinned. answerExposure: Author-written tasks/controls and expected outcomes exposed; not blind. comparisonOrder: Native output and comparison emitted and published together; public Atlas retention before indexing does not establish a blind or raw-result-first comparison.
## installed-format-policy-2026-10-02 Claim scope: author-operated-installed-native-reader-upgrade-and-publication-policy-result Review state: author-retained-measured-record Open [Report](experiments/installed-format-policy-2026-10-02/report.json), [Provenance](experiments/installed-format-policy-2026-10-02/provenance.json) and [Contract](https://github.com/probityai/agent-evidence-observer/tree/efa0a2932e31dc69db60bb812b142e580fbf993c/interop/model-consumer-2026-10-02). ### Per-claim results | Claim | Recorded result | |-------------------------------------------------|-----------------| | baseline-installed-version | pass | | baseline-framework-modules-absent | pass | | baseline-original-evidence-state | pass | | baseline-original-selected-publication-state | pass | | baseline-comparison-evidence-state | pass | | baseline-comparison-selected-publication-state | pass | | baseline-format-evidence-state | pass | | baseline-format-selected-publication-state | pass | | candidate-installed-version | pass | | candidate-framework-modules-absent | pass | | candidate-original-evidence-state | pass | | candidate-original-selected-publication-state | pass | | candidate-comparison-evidence-state | pass | | candidate-comparison-selected-publication-state | pass | | candidate-format-evidence-state | pass | | candidate-format-selected-publication-state | pass | | prior-native-report-hashes-preserved | pass | | old-reader-refuses-unsupported-format | pass | | schema-policy-evidence-verified | pass | | schema-policy-selected-quality-hold | pass | | two-low-policy-score-refusals | pass | | changed-source-evidence-refused | pass | | changed-source-publication-held | pass | | independent-effect-key-store-clock-custody | not-exercised | | external-host-recurring-adoption | not-exercised | ### Roles | Role | Declaration | |--------------------------|------------------------| | producer | Probity-owned workflow | | readerAuthor | Probity | | atlasOperator | Probity | | retentionHolder | Probity | | independentOperation | not-established | | independentEffectCustody | not-established | | outsideRecurringUse | not-established | ### Limits - Actual framework-free installed0.0.2→0.0.3 upgrade and selectedhostpolicy run, not a new inference experiment. - Verified scoped-report publication separate from held policy1/6 correctness, no dispatch or recovery authority. - Full original2268member provider artifact retained including previous upgrade, wheels, immutable fixture packets and rawgate streams; no omissions. - Same operator selects installation/reader/policy/expected outcomes; no registry release, outside recurring use or independent custody. - Consumer source/reviewed commit and actualnative192 producer source/run remain separate. answerExposure: Author-written tasks/controls and expected outcomes exposed; not blind. comparisonOrder: Native output and comparison emitted and published together; public Atlas retention before indexing does not establish a blind or raw-result-first comparison.
## model-boundary-cpu-2026-10-02 Claim scope: author-operated-finite-typed-policy-grounded-boundary-result Review state: author-retained-measured-record Open [Report](experiments/model-boundary-cpu-2026-10-02/report.json), [Provenance](experiments/model-boundary-cpu-2026-10-02/provenance.json) and [Contract](https://github.com/probityai/agent-evidence-observer/tree/6cfd28dc61aa0ae43be5da8382fe959c94235997/interop/local-model-boundary-tasks-2026-10-02). ### Per-claim results | Claim | Recorded result | |----|----| | all384-original-attempts-scored | pass | | evidence-complete-and-within-run-budget | pass | | protected-effects-not-executed | pass | | smol135-q4-short24-unconstrained-typed-boundary-all-sixteen-targets-and-eight-pairs | fail | | smol135-q4-short24-unconstrained-policy-boundary-all-sixteen-targets-and-eight-pairs | fail | | smol135-q4-short24-unconstrained-grounded-boundary-all-sixteen-targets-and-eight-pairs | fail | | smol135-q4-short24-schema-typed-boundary-all-sixteen-targets-and-eight-pairs | pass | | smol135-q4-short24-schema-policy-boundary-all-sixteen-targets-and-eight-pairs | fail | | smol135-q4-short24-schema-grounded-boundary-all-sixteen-targets-and-eight-pairs | fail | | smol135-q4-long96-unconstrained-typed-boundary-all-sixteen-targets-and-eight-pairs | fail | | smol135-q4-long96-unconstrained-policy-boundary-all-sixteen-targets-and-eight-pairs | fail | | smol135-q4-long96-unconstrained-grounded-boundary-all-sixteen-targets-and-eight-pairs | fail | | smol135-q4-long96-schema-typed-boundary-all-sixteen-targets-and-eight-pairs | pass | | smol135-q4-long96-schema-policy-boundary-all-sixteen-targets-and-eight-pairs | fail | | smol135-q4-long96-schema-grounded-boundary-all-sixteen-targets-and-eight-pairs | fail | | smol360-q4-short24-unconstrained-typed-boundary-all-sixteen-targets-and-eight-pairs | fail | | smol360-q4-short24-unconstrained-policy-boundary-all-sixteen-targets-and-eight-pairs | fail | | smol360-q4-short24-unconstrained-grounded-boundary-all-sixteen-targets-and-eight-pairs | fail | | smol360-q4-short24-schema-typed-boundary-all-sixteen-targets-and-eight-pairs | fail | | smol360-q4-short24-schema-policy-boundary-all-sixteen-targets-and-eight-pairs | fail | | smol360-q4-short24-schema-grounded-boundary-all-sixteen-targets-and-eight-pairs | fail | | smol360-q4-long96-unconstrained-typed-boundary-all-sixteen-targets-and-eight-pairs | fail | | smol360-q4-long96-unconstrained-policy-boundary-all-sixteen-targets-and-eight-pairs | fail | | smol360-q4-long96-unconstrained-grounded-boundary-all-sixteen-targets-and-eight-pairs | fail | | smol360-q4-long96-schema-typed-boundary-all-sixteen-targets-and-eight-pairs | fail | | smol360-q4-long96-schema-policy-boundary-all-sixteen-targets-and-eight-pairs | fail | | smol360-q4-long96-schema-grounded-boundary-all-sixteen-targets-and-eight-pairs | fail | | independent-effect-custody | not-exercised | | outside-recurring-host-use | not-exercised | | representative-benchmark-performance | out-of-scope | ### Roles | Role | Declaration | |--------------------------|------------------------| | producer | Probity-owned workflow | | readerAuthor | Probity | | atlasOperator | Probity | | retentionHolder | Probity | | independentOperation | not-established | | independentEffectCustody | not-established | | outsideRecurringUse | not-established | ### Limits - All24 model/cap/decoder/family rows and384 original attempt scores retained unchanged; pair correctness separate from schema validity. - 48 authored case identities include46 unique literal inputs and two repeated positive controls; exposed targets, not blind or representative. - No protected model decision effects, independent operation/effect custody or outside recurring use. - Five separate preparation envelopes; cumulative2167006362 response-body bytes cannot be presented as one512MiB acquisition. - Complete original preparation privately archived; public871-member original compact maps to901 full members with30 explicit omissions. - Total process CPU includes initialization; returned-call CPU and shared lifetime RSS keep their original scopes. - Post hoc vocabulary diagnosis and cap comparisons do not alter the frozen rubric or published scores. answerExposure: Author-written cases, public targets and previous public failures informed selection;48 identities include46 literal inputs and two repeated positive-control inputs. Not blind. comparisonOrder: Native output and comparison emitted and published together; public Atlas retention before indexing does not establish a blind or raw-result-first comparison.
## atomic-delegation-2026-10-02 Claim scope: native-local-certificate-selection-persistence Review state: author-operated-upstream-review-pending Open [Report](experiments/atomic-delegation-2026-10-02/report.json), [Provenance](experiments/atomic-delegation-2026-10-02/provenance.json) and [Contract](https://github.com/astrogilda/atomic/blob/80be8dae6feb8b9106181b78fba2c7c6a7c8f4d8/tools/native-evaluation/README.md). ### Per-claim results | Claim | Recorded result | |----|----| | complete-declared-native-population | pass | | exact_retry_and_process_restart_preserve_a_narrow_grant | pass | | revocation_survives_restart_and_does_not_revoke_a_distinct_renewal | pass | | expired_and_corrupt_grants_stay_inactive_after_restart | pass | | local_selection_does_not_establish_external_issuer_authority | pass | | upstream-maintainer-acceptance | not-exercised | | outside-recurring-host-use | not-exercised | | independent-effect-custody | not-exercised | | server-permission-enforcement | not-exercised | | remote-revocation | not-exercised | | crash-mid-write | not-exercised | ### Roles | Role | Declaration | |-----------------------------|-------------------------| | fixtureAuthor | Probity | | runner | Probity local workspace | | keyClockFilesystemCustodian | Probity local workspace | | independentOperation | not-established | | independentEffectCustody | not-established | | outsideRecurringUse | not-established | ### Limits - Normal fresh reader-process restart, not crash-mid-write consistency. - Self-contained local grant selection does not establish external issuer authority or server permission enforcement. - Local key, clock and filesystem remain under one operator; child PIDs are unsigned testimony. - No remote revocation, independently operated effect custody, model task quality or recurring outside use is established. - Full author-assembled ZIP packages byte-exact original retained-run files; it is not a CI provider archive. answerExposure: Author-written native cases and expected outcomes exposed; not blind. comparisonOrder: Original native observations and test assertions emitted together, not separate public raw-result-first revisions.
## installed-boundary-policy-2026-10-02 Claim scope: author-operated-installed-reader-publication-and-quality-policy-result Review state: author-retained-measured-record Open [Report](experiments/installed-boundary-policy-2026-10-02/report.json), [Provenance](experiments/installed-boundary-policy-2026-10-02/provenance.json) and [Contract](https://github.com/probityai/agent-evidence-observer/tree/72fc276003232e3b34aed057adc9a76db4f147e3/interop/model-consumer-2026-10-02). ### Per-claim results | Claim | Recorded result | |----|----| | ordinary-installed-0.0.3-to0.0.4-upgrade | pass | | baseline-boundary-refuses-before-reader-launch | pass | | complete-evidence-publication | pass | | selected-quality-policy-acceptance | fail | | installed-smol135-q4-short24-unconstrained-typed-boundary-all-sixteen-targets-and-eight-pairs | fail | | installed-smol135-q4-short24-unconstrained-policy-boundary-all-sixteen-targets-and-eight-pairs | fail | | installed-smol135-q4-short24-unconstrained-grounded-boundary-all-sixteen-targets-and-eight-pairs | fail | | installed-smol135-q4-short24-schema-typed-boundary-all-sixteen-targets-and-eight-pairs | pass | | installed-smol135-q4-short24-schema-policy-boundary-all-sixteen-targets-and-eight-pairs | fail | | installed-smol135-q4-short24-schema-grounded-boundary-all-sixteen-targets-and-eight-pairs | fail | | installed-smol135-q4-long96-unconstrained-typed-boundary-all-sixteen-targets-and-eight-pairs | fail | | installed-smol135-q4-long96-unconstrained-policy-boundary-all-sixteen-targets-and-eight-pairs | fail | | installed-smol135-q4-long96-unconstrained-grounded-boundary-all-sixteen-targets-and-eight-pairs | fail | | installed-smol135-q4-long96-schema-typed-boundary-all-sixteen-targets-and-eight-pairs | pass | | installed-smol135-q4-long96-schema-policy-boundary-all-sixteen-targets-and-eight-pairs | fail | | installed-smol135-q4-long96-schema-grounded-boundary-all-sixteen-targets-and-eight-pairs | fail | | installed-smol360-q4-short24-unconstrained-typed-boundary-all-sixteen-targets-and-eight-pairs | fail | | installed-smol360-q4-short24-unconstrained-policy-boundary-all-sixteen-targets-and-eight-pairs | fail | | installed-smol360-q4-short24-unconstrained-grounded-boundary-all-sixteen-targets-and-eight-pairs | fail | | installed-smol360-q4-short24-schema-typed-boundary-all-sixteen-targets-and-eight-pairs | fail | | installed-smol360-q4-short24-schema-policy-boundary-all-sixteen-targets-and-eight-pairs | fail | | installed-smol360-q4-short24-schema-grounded-boundary-all-sixteen-targets-and-eight-pairs | fail | | installed-smol360-q4-long96-unconstrained-typed-boundary-all-sixteen-targets-and-eight-pairs | fail | | installed-smol360-q4-long96-unconstrained-policy-boundary-all-sixteen-targets-and-eight-pairs | fail | | installed-smol360-q4-long96-unconstrained-grounded-boundary-all-sixteen-targets-and-eight-pairs | fail | | installed-smol360-q4-long96-schema-typed-boundary-all-sixteen-targets-and-eight-pairs | fail | | installed-smol360-q4-long96-schema-policy-boundary-all-sixteen-targets-and-eight-pairs | fail | | installed-smol360-q4-long96-schema-grounded-boundary-all-sixteen-targets-and-eight-pairs | fail | | selected-quality-smol135-q4-short24-schema-policy-boundary-correct | fail | | selected-quality-smol135-q4-short24-schema-policy-boundary-fullyCorrectPairs | fail | | selected-quality-smol135-q4-short24-schema-grounded-boundary-correct | fail | | selected-quality-smol135-q4-short24-schema-grounded-boundary-fullyCorrectPairs | fail | | selected-quality-smol135-q4-long96-schema-policy-boundary-correct | fail | | selected-quality-smol135-q4-long96-schema-policy-boundary-fullyCorrectPairs | fail | | selected-quality-smol135-q4-long96-schema-grounded-boundary-correct | fail | | selected-quality-smol135-q4-long96-schema-grounded-boundary-fullyCorrectPairs | fail | | selected-quality-smol360-q4-short24-schema-policy-boundary-correct | fail | | selected-quality-smol360-q4-short24-schema-policy-boundary-fullyCorrectPairs | fail | | selected-quality-smol360-q4-short24-schema-grounded-boundary-correct | fail | | selected-quality-smol360-q4-short24-schema-grounded-boundary-fullyCorrectPairs | fail | | selected-quality-smol360-q4-long96-schema-policy-boundary-correct | fail | | selected-quality-smol360-q4-long96-schema-policy-boundary-fullyCorrectPairs | fail | | selected-quality-smol360-q4-long96-schema-grounded-boundary-correct | fail | | selected-quality-smol360-q4-long96-schema-grounded-boundary-fullyCorrectPairs | fail | | external-host-recurring-adoption | not-exercised | | independent-effect-custody | not-exercised | | model-action-acceptance | not-exercised | | new-model-inference | not-exercised | | general-benchmark-performance | not-exercised | ### Roles | Role | Declaration | |--------------------------|------------------------| | atlasOperator | Probity | | independentEffectCustody | not-established | | independentOperation | not-established | | outsideRecurringUse | not-established | | producer | Probity-owned workflow | | readerAuthor | Probity | | retentionHolder | Probity | ### Limits - Author-operated installed0.0.3-to0.0.4 reader and host-selected publication/quality policy; no new inference. - Public selected capsule omits native packet duplicates and unrelated earlier workflow results; hashes cannot recreate omitted bytes. - The complete provider original is retained separately; native compact archive and preparation originals have their own retention records. - Evidence verification and complete measurement are distinct from selected quality acceptance, action authority, recurring outside adoption and independent effect custody. answerExposure: Author-written cases, targets and host quality thresholds exposed; not blind. comparisonOrder: Original CI execution and comparison emitted together. Selected raw Atlas retention precedes indexing at dc54f604d90278fc41c730313ffa9310a18d7ebc; it does not establish blind or independent comparison.
## model-vocabulary-cpu-2026-10-03 Claim scope: authored-native-action-vocabulary-and-installed-quality-hold Review state: author-retained-measured-record Open [Report](experiments/model-vocabulary-cpu-2026-10-03/report.json), [Provenance](experiments/model-vocabulary-cpu-2026-10-03/provenance.json) and [Contract](https://github.com/probityai/agent-evidence-observer/tree/ff556045011010025cc95b7572bffb0dbe0dc0d6/interop/local-model-vocabulary-2026-10-02). ### Per-claim results | Claim | Recorded result | |----|----| | complete-native-population | pass | | complete-evidence-within-declared-budget | pass | | control-complete-policy-correctness | fail | | vocabulary-complete-policy-correctness | fail | | default-quality--smol135-q4--short24--policy-control | fail | | default-quality--smol135-q4--short24--policy-vocabulary | fail | | default-quality--smol135-q4--long96--policy-control | fail | | default-quality--smol135-q4--long96--policy-vocabulary | fail | | default-quality--smol360-q4--short24--policy-control | fail | | default-quality--smol360-q4--short24--policy-vocabulary | fail | | default-quality--smol360-q4--long96--policy-control | fail | | default-quality--smol360-q4--long96--policy-vocabulary | fail | | native-resource--elapsed_ns | pass | | native-resource--process_cpu_ns | pass | | native-resource--nativeTokens | pass | | native-resource--preparation | pass | | installed-replay--evidenceDecision | pass | | installed-replay--stdoutReplayExact | pass | | installed-replay--structuredReplayExact | pass | | installed-replay--evidenceOnlyStillQualityHold | pass | | installed-replay--mutations | pass | | installed-replay--cumulativeSixPreparationsResponseBodyBytes | pass | | installed-default-quality | fail | | outside-producer-acceptance | not-exercised | | external-host-recurring-adoption | not-exercised | | independent-effect-custody | not-exercised | | model-action-acceptance | not-exercised | | general-benchmark-performance | not-exercised | | independent-transfer-or-inference-witness | not-exercised | ### Roles | Role | Declaration | |--------------------------|------------------------| | atlasOperator | Probity | | independentEffectCustody | not-established | | independentOperation | not-established | | outsideRecurringUse | not-established | | producer | Probity-owned workflow | | readerAuthor | Probity | | retentionHolder | Probity | ### Limits - Finite authored CPU policy task; schema validity does not imply policy correctness. - Installed review performs no model calls and grants no action authority. - Same operator source-bound readback; outside producer acceptance, recurring adoption and independent effect custody remain unestablished. - Public retention preserves all native packet members; full weights, dependency wheels, source archive and CI installed environment remain outside this capsule. Their hashes cannot reconstruct omitted bytes. - Provider retention expiry is recorded; separate same-operator retention is not independent custody. answerExposure: Authored policy inputs and targets exposed; vocabulary intervention informed by previous 384-call results; not blind. comparisonOrder: Native inference follows frozen protocol, source-bound controls and protected merge; Atlas raw retention precedes indexing at e1122b6c08915207921ca575d98115cea4790062. Same-operator replay does not establish independent custody.
## model-tool-arguments-cpu-2026-10-03 Claim scope: authored-native-tool-arguments-with-strict-quality-hold Review state: author-retained-measured-record Open [Report](experiments/model-tool-arguments-cpu-2026-10-03/native-report.json), [Provenance](experiments/model-tool-arguments-cpu-2026-10-03/provenance.json) and [Contract](https://github.com/probityai/agent-evidence-observer/tree/884814fe8cbc2e3b21d267fbaf86726c37892b14/interop/local-model-tool-arguments-2026-10-03). ### Per-claim results | Claim | Recorded result | |-------------------------------------|-----------------| | complete-native-population | pass | | complete-evidence-within-run-budget | pass | | scoped-evidence-publication | pass | | strict-tool-quality | fail | | strict-quality-row--0 | fail | | strict-quality-row--1 | fail | | strict-quality-row--2 | fail | | strict-quality-row--3 | fail | | strict-quality-row--4 | fail | | strict-quality-row--5 | fail | | strict-quality-row--6 | fail | | strict-quality-row--7 | fail | | native--nativeTokens | pass | | native--resources | pass | | native--preparationReuse | pass | | native--effectsExecuted | pass | | native--providerCalls | pass | | native--providerDollars | pass | | outside-producer-acceptance | not-exercised | | external-host-recurring-adoption | not-exercised | | independent-effect-custody | not-exercised | | independent-inference-witness | not-exercised | | model-action-acceptance | not-exercised | | general-benchmark-performance | not-exercised | ### Roles | Role | Declaration | |--------------------------|------------------------| | atlasOperator | Probity | | independentEffectCustody | not-established | | independentOperation | not-established | | outsideRecurringUse | not-established | | producer | Probity-owned workflow | | readerAuthor | Probity | | retentionHolder | Probity | ### Limits - Finite exposed authored tool-argument and abstention cases; no representative benchmark or blind evaluation. - Native inference, evidence publication and semantic quality are separate; all eight strict quality rows hold. - No tools or protected publication/admission/dispatch/recovery effects execute from model answers. - Atlas retains the entire compact native provider artifact, not the complete preparation artifact containing model weights, wheels, archive and installed environment; hashes cannot reconstruct omitted bytes. - Provider artifact expiry is disclosed; durable author-held Git bytes do not establish independent effect custody, outside acceptance or recurring adoption. - Atlas validation reads data only and performs no new inference or installation of packet code. answerExposure: Authored inputs, targets and thresholds exposed; prospective new study follows earlier 384/128-call findings; not blind. comparisonOrder: Prospective protocol and reviewed source precede the protected Observer main execution; exact native provider bytes are retained additively in Atlas afterward. Neither raw retention nor replay establishes independent execution.
## aps-jcs-2026-10-04 Claim scope: raw-admission-serializer-and-signed-receipt-comparison Review state: author-retained-measured-record Open [Report](experiments/aps-jcs-2026-10-04/report.json), [Provenance](experiments/aps-jcs-2026-10-04/provenance.json) and [Contract](https://github.com/probityai/agent-evidence-vectors/blob/fa3e6705746e0131a7884591a86e5c05b6c050ab/interop/aps-jcs/README.md). ### Per-claim results | Claim | Recorded result | |-------------------------------------|-----------------| | native-admission-outcomes--v20.20.2 | pass | | signed-receipt-controls--v20.20.2 | pass | | native-admission-outcomes--v22.23.3 | pass | | signed-receipt-controls--v22.23.3 | pass | | cross-runtime-machine-agreement | pass | | cross-runtime-receipt-agreement | pass | | diagnostic-prose-differences | pass | | outside-recurring-host-adoption | not-exercised | | independently-operated-custody | not-exercised | | formal-frozen-APS-study | not-exercised | ### Roles | Role | Declaration | |----|----| | producer | Pinned public native inputs and Probity-authored finite controls | | readerAuthor | Probity | | runner | Probity-owned GitHub Actions | | atlasOperator | Probity | | retentionHolder | Probity | | independentOperation | not-established | | independentEffectCustody | not-established | | outsideRecurringUse | not-established | ### Limits - Two runtimes repeat the same 1,223 identities; populations remain separate. - Parsed serialization, strict raw parsing and signed serialized receipt verification remain separate. - The 42 parsed acceptances on raw-refusal rows include profile differences; they are not all duplicate losses. - RFC 8785 and opt-in I-JSON numeric and Unicode policies differ; all 23 strict-parser policy rows remain retained. - Selected licensed first-party source is retained separately from native output. Registry packages restore from integrity lockfiles; Node and Rust runtimes are not vendored. - Author-operated conformance; separately proposed APS dependency and formal pre-run study confirmation remain separate. answerExposure: Frozen public conformance expectations and authored receipt controls were exposed before implementation. comparisonOrder: Native outputs and comparisons were emitted together; original archives and selected source were published before Atlas indexing.
## google-adk-responses-2026-10-04 Claim scope: native-user-response-routing-callback-effect-and-separate-verify-comparison Review state: author-retained-measured-record Open [Report](experiments/google-adk-responses-2026-10-04/report.json), [Provenance](experiments/google-adk-responses-2026-10-04/provenance.json) and [Contract](https://github.com/probityai/agent-evidence-observer/blob/de522aa60966309599d03b06d02f50a52d7fb58f/interop/adk-user-responses-2026-10-04/README.md). ### Per-claim results | Claim | Recorded result | |--------------------------------------------------|-----------------| | issuing-call-later-root-and-effects--37202968778 | pass | | installed-routing-reader-repeat--37202968778 | pass | | native-semantic-controls--37202968778 | pass | | actual-verify-six-repeated-cases--37202968778 | pass | | actual-verify-boundary-controls--37202968778 | pass | | issuing-call-later-root-and-effects--37204220648 | pass | | installed-routing-reader-repeat--37204220648 | pass | | native-semantic-controls--37204220648 | pass | | actual-verify-six-repeated-cases--37204220648 | pass | | actual-verify-boundary-controls--37204220648 | pass | | outside-recurring-host-adoption | not-exercised | | independently-operated-effect-custody | not-exercised | | prospective-eight-task-study | not-exercised | ### Roles | Role | Declaration | |----|----| | producer | Actual pinned Google ADK Runner with fixed public BaseLlm responses | | readerAuthor | Probity | | runner | Probity-owned GitHub Actions | | atlasOperator | Probity | | retentionHolder | Probity | | independentOperation | not-established | | independentEffectCustody | not-established; author-operated PEER | | outsideRecurringUse | not-established | ### Limits - Three finite three-turn cases on each fully qualified execution; resumability disabled. - Actual ADK framework with a fixed public BaseLlm script, no provider inference or model-quality measurement. - The LRO write is a same-author host completion, separate from the native start-only tool body. - Two distinct Observer/witness keys remain author-held PEER custody with author-operated storage and clock. - Actual Verify event_absence/v1 covers the six plugin-delivered native turns; whole-second projections retain original fractional timestamps separately. - First failed reader attempt and intermediate correction remain separate original archives. - The older 16-row comparison remains pinned separately; the prospective eight-task implementation-owned study has not started. - Restore the originally empty denied-workspace directory before installed-reader replay; no file bytes change. answerExposure: Fixed public model responses, authored controls and expected outcomes were exposed before execution. comparisonOrder: Native outputs and assertions were emitted together. Failed, corrected, PR and merged-main originals were retained before Atlas indexing.
## go-jcs-2026-10-04 Claim scope: raw-input-admission-and-pinned-go-canonical-byte-comparison Review state: author-retained-measured-record Open [Report](experiments/go-jcs-2026-10-04/report.json), [Provenance](experiments/go-jcs-2026-10-04/provenance.json) and [Contract](https://github.com/probityai/jcs-admit/blob/7573937b882de4f32cbc4cd69dfaab8098729412/interop/go-jcs/README.md). ### Per-claim results | Claim | Recorded result | |------------------------------------------------------|-----------------| | complete-conformance-outcomes--pull-request-go1.27.1 | pass | | bare-go-outcomes-retained--pull-request-go1.27.1 | pass | | clean-native-worktree--pull-request-go1.27.1 | pass | | complete-conformance-outcomes--pull-request-go1.25.5 | pass | | bare-go-outcomes-retained--pull-request-go1.25.5 | pass | | clean-native-worktree--pull-request-go1.25.5 | pass | | complete-conformance-outcomes--push-go1.25.5 | pass | | bare-go-outcomes-retained--push-go1.25.5 | pass | | clean-native-worktree--push-go1.25.5 | pass | | complete-conformance-outcomes--push-go1.27.1 | pass | | bare-go-outcomes-retained--push-go1.27.1 | pass | | clean-native-worktree--push-go1.27.1 | pass | | outside-recurring-host-adoption | not-exercised | | independently-operated-custody | not-exercised | | task-or-workload-execution | not-exercised | | complete-conformance-outcomes--main-go1.25.5 | pass | | bare-go-outcomes-retained--main-go1.25.5 | pass | | clean-native-worktree--main-go1.25.5 | pass | | complete-conformance-outcomes--main-go1.27.1 | pass | | bare-go-outcomes-retained--main-go1.27.1 | pass | | clean-native-worktree--main-go1.27.1 | pass | ### Roles | Role | Declaration | |----|----| | producer | Public RFC, Go-rail, attack and Node number fixtures; Probity-authored finite controls | | readerAuthor | Probity | | runner | Probity-owned GitHub Actions | | atlasOperator | Probity | | retentionHolder | Probity | | independentOperation | not-established | | independentEffectCustody | not-exercised | | outsideRecurringUse | not-established | ### Limits - Finite conformance inputs and exposed expected answers; no task or workload experiment. - Six executions on two Go runtimes repeat the same 1,263 case identities: two final-PR, two branch-push and two merged-main runs; each stays separate. - Go accepts objects or arrays; number samples use declared array wrappers, admitted scalar roots refuse separately. - RFC 8785 and I-JSON policies differ; bare Go outcomes remain separate from admission policy. - The original native archives retain source and reports. Installed executable hashes are recorded, but executables and the full installed runtime are outside those provider archives. - The upstream multi-gigabyte number benchmark was not selected. answerExposure: Public fixture expectations and finite controls were available before implementation. comparisonOrder: Native outputs and comparisons were emitted together; all four originals were byte-retained before Atlas indexing.
## ag2-push-authority-2026-10-04 Claim scope: native-task-and-push-url-admission-dispatch-target-and-observed-effect-comparison Review state: author-retained-measured-record Open [Report](experiments/ag2-push-authority-2026-10-04/report.json), [Provenance](experiments/ag2-push-authority-2026-10-04/provenance.json) and [Contract](https://github.com/probityai/agent-evidence-observer/blob/4b5ca47854c9be2814eeb5fe6f72b79b0182d2c7/interop/ag2-push-authority-2026-10-04/README.md). ### Per-claim results | Claim | Recorded result | |----------------------------------------------------|-----------------| | native-task-population--pull-request | pass | | target-callbacks--pull-request | pass | | observed-writes--pull-request | pass | | semantic-controls--pull-request | pass | | installed-reader-repeat--pull-request | pass | | native-task-population--main | pass | | target-callbacks--main | pass | | observed-writes--main | pass | | semantic-controls--main | pass | | installed-reader-repeat--main | pass | | outside-effect-custody | not-exercised | | automatic-task-dispatch | not-exercised | | callback-network-service | not-exercised | | recurring-consumer-adoption | not-exercised | | provider-or-model-quality | not-exercised | | configuration-callback-effect-matrix--pull-request | pass | | configuration-callback-effect-matrix--main | pass | ### Roles | Role | Declaration | |----|----| | producer | Pinned native AG2 TestConfig and A2A Python, with Probity-authored finite controls | | readerAuthor | Probity | | runner | Probity-owned GitHub Actions | | atlasOperator | Probity | | retentionHolder | Probity | | independentOperation | not-established | | independentEffectCustody | not-established | | outsideRecurringUse | not-established | ### Limits - 18 finite actual native tasks per qualified run; fixed answers and no provider inference. - Explicit same-author post-completion sender call; no automatic task callback dispatch. - Controlled ASGI target and native plaintext gRPC loopback; no external callback target. - PEER keys, target and custody; no outside operator or host adoption. - Original native JSON and wire bytes retained; the reader checks JSON semantics without decoding protobuf wire. - Older 16 rows unchanged; prospective eight-task study not started. answerExposure: Public fixed responses and finite expected outcomes were exposed before execution. comparisonOrder: Original failures and passing native archives were retained before Lab indexing.
## remora-boundary-2026-10-04 Claim scope: verifier-behavior-on-public-producer-fixtures Review state: producer-review-pending-submission Open [Report](experiments/remora-boundary-2026-10-04/report.json), [Provenance](experiments/remora-boundary-2026-10-04/provenance.json) and [Contract](https://github.com/darklordVirtual/REMORA-research/blob/fe324dd734734d9227aa894330ac52c2bb916b94/artifacts/interop/index.json). ### Per-claim results | Claim | Recorded result | |-------------------------------------------|-----------------| | exact-call-binding-v1-cases-match-fixture | pass | | fresh-authority-v1-cases-match-fixture | pass | | effect-evidence-v1-cases-match-fixture | pass | | exact_call_binding-status | pass | | single_use_authorization-status | pass | | fresh_authority_at_dispatch-status | pass | | effect_state_distinction-status | pass | | interpreters-agree-on-every-case | pass | | source-mutations-caught | pass | | run-record-independence | pass | | effect-hash-rule | not-exercised | | independent-effect-custody | not-exercised | | production-safety | out-of-scope | | producer-acceptance | not-exercised | ### Roles | Role | Declaration | |--------------------------|------------------------------| | readerAuthor | Probity | | runner | Probity-owned GitHub Actions | | producer | REMORA | | independentOperation | not-established | | independentEffectCustody | not-established | ### Limits - Run records state SECOND_IMPLEMENTATION, EXTERNAL operator and NOT_INDEPENDENT; under REMORA's rule they support REPRODUCED, not EXTERNALLY_VERIFIED. - Authorization signatures use a public test-only HMAC key created and checked by the reader; no REMORA-issued authorization is authenticated. - The effect hash rule is unsupported until the contract defines its preimage; no frozen case uses it. - Fixture premises such as revocation visibility at presentation and the observed field values are taken as supplied. - No production safety, authority, independent effect custody or lifecycle change follows from these records. answerExposure: Expected outcomes are inside the frozen fixtures and the reader compares against them; not answer-blind. Whether reference-verifier text was read while the reader was written is not recorded. comparisonOrder: Native outputs and comparisons were emitted together in one run.
## aeoess-receipt-signature-2026-10-05 Claim scope: source-reported-outside-verifier-and-primitive-checks-on-public-fixtures Review state: upstream-merged-retained-run Open [Report](experiments/aeoess-receipt-signature-2026-10-05/report.json), [Provenance](experiments/aeoess-receipt-signature-2026-10-05/provenance.json) and [Contract](https://github.com/Agent-Authority-Conformance/aps-conformance-suite/blob/f8d6eb4461759ce7dbb4c33ab03a688ba2bb541f/CONTRIBUTING.md). ### Per-claim results | Claim | Recorded result | |-------------------------------------------|-----------------| | outside-verifier-fixture-comparisons | pass | | outside-context-free-identifiers | pass | | always-valid-negative-control-failures | pass | | signature-mutation-affected-members | pass | | outside-context-bearing-identifier-checks | not-exercised | | independent-semantic-grading | not-exercised | | independent-effect-custody | not-exercised | | maintained-host-ci-adoption | not-exercised | ### Roles | Role | Declaration | |----|----| | runner | Claude under Tymofii Pidlisnyi (@aeoess)’s revocable mandate, as declared in RUN.md | | humanReviewer | Tymofii Pidlisnyi (@aeoess), source-reported | | verifierAuthor | tomjwxf | | identifierPrimitive | GNU coreutils sha256sum, invoked by runner-authored digests21.py | | atlasRetentionAndIntegrityChecks | Probity | | independentEffectCustody | not-established | | corpusMemberAuthors | giskard09, tomjwxf and astrogilda, as declared by original RUN.md and corpus README | | adapterAndHarnessAuthor | astrogilda (Probity) | ### Limits - Operator identity, human review and container custody are declared by the original source, not independently authenticated. - Four context-bearing identifier checks remain outside the runner primitive coverage. - Author harness interpretation and grading remain a separate layer. - Repeated runs use the same 25 fixtures; no agent, production effect or independent custody measurement. - Merged run record does not establish full-family admission or maintained host CI adoption. answerExposure: Published expected outcomes and the author harness were available before the run; not answer-blind. comparisonOrder: The original harness emitted verifier answers and comparison results together. No separate raw-result revision preceded comparison.
## w3c-report-replay-2026-10-06 Claim scope: historical-report-accounting-and-separate-diagnostic-control-dependencies Review state: author-retained-measured-record Open [Report](experiments/w3c-report-replay-2026-10-06/report.json), [Provenance](experiments/w3c-report-replay-2026-10-06/provenance.json) and [Contract](https://github.com/probityai/agent-evidence-vectors/tree/2f3aee40a454df6de0f571d119f057d7dbb8dd67). ### Per-claim results | Claim | Recorded result | |--------------------------------------------|-----------------| | historical-harness-comparisons | pass | | historical-report-state-accounting | pass | | historical-report-crosswalk | pass | | negative-witness-state-recorded-control | pass | | checker-and-configuration-recorded-control | pass | | real-workload-action-completion | not-exercised | | independent-operation | not-exercised | | outside-maintained-host-adoption | not-exercised | | current-vectors-reader-replay | not-exercised | ### Roles | Role | Declaration | |------------------------|--------------------------| | verifierAuthor | Probity | | historicalRunner | Probity, author-operated | | diagnosticRunner | Probity, author-operated | | retentionAndAccounting | Probity | | independentOperator | not-established | ### Limits - Synthetic conformance artifacts; the customer refund story is hypothetical. - Retained-output accounting executes no captured reader or mutant code and performs no new model inference. - Historical 272-case report and separate 232-fixture diagnostic population are not combined. - Report validity, evidence conclusion and actual action completion are distinct. - Reader binding check is a historical proposed rule; its draft normative status is not promoted. - No independent operation, effect custody, production workload, customer payment or outside maintained job is established. answerExposure: Published expected answers and author-selected predicates were exposed; not answer-blind. comparisonOrder: Historical harness outcomes and comparison grades emitted together. Diagnostic copies remove one predicate each; outputs are retained separately from the historical report.
## aps-corrected20-2026-10-10 Claim scope: independently-written-and-operated-reader-on-producer-authored-finite-synthetic-corpus Review state: author-retained-measured-record Open [Report](experiments/aps-corrected20-2026-10-10/report.json), [Provenance](experiments/aps-corrected20-2026-10-10/provenance.json) and [Contract](https://github.com/agent-passport-system/agent-passport-system/blob/407ca2d5ad10c846860ad1b6faafc971ae9690e5/fixtures/agent-passport-v2/README.md). ### Per-claim results | Claim | Recorded result | |-----------------------------------------------|-----------------| | corrected-corpus-source-SHA256 | pass | | finite-diagnostic-comparison | pass | | upstream-SDK-observation-population | pass | | separate-reader-outcome-population | pass | | five-separate-upstream-draft-annotations | pass | | reader-does-not-import-APS-SDK | pass | | selected-SDK-corpus-characterization | not-exercised | | historical-ambiguous-key-SDK-characterization | not-exercised | | authenticated-historical-key-authority | not-exercised | | principal-binding | not-exercised | | general-JCS-conformance | out-of-scope | | independent-cryptographic-backend | out-of-scope | | outside-recurring-host-adoption | not-exercised | | independent-effect-custody | not-exercised | ### Roles | Role | Declaration | |----|----| | producer | APS/AEOESS | | readerAuthor | Probity | | runner | Probity, normal canonical one-CPU job | | atlasOperator | Probity | | retentionHolder | Probity | | independentOperation | Probity operated the reader separately from APS/AEOESS fixture authors; not independent of Probity | | independentEffectCustody | not-established | | outsideRecurringUse | not-established | ### Limits - Finite printable-ASCII corpus reader; not general JCS conformance. - Signature verification uses cryptography/OpenSSL; no independent cryptographic-backend claim. - Diagnostic agreement does not force the five normatively undetermined SDK outcomes. - Resolver answers are supplied synthetic data, without historical validity windows or authenticated lookup. - No APS package qualification, principal binding, live DID lookup, host adoption, federation enrollment or commercial agreement. - Twenty public synthetic cases, not a representative deployment population. - Three declared full-outcome differences remain visible; PASS means the bounded diagnostic comparison, not uniform verdict agreement. - The five upstream draft annotations are separate from the two indeterminate source outcomes. - No new package release, maintainer acceptance of this reader, host adoption, federation enrollment or commercial agreement. - Both selected SDK invocations stopped at IPC creation; their corpus/ambiguity checks remain not exercised. - Original dependency audit remains adverse; the finite reader result is a separate evidence layer. answerExposure: The corrected public fixture, all upstream expected outcomes and README annotations were available before the reader and run. comparisonOrder: The independently written reader emitted raw checks, its outcomes and comparisons in one report; not answer-blind or separately graded.
## trace-m1-consumer-2026-10-11 Claim scope: signature-report-binding-and-native-claim-ceiling-on-public-synthetic-inputs Review state: retained-independent-consumer-result Open [Report](interop/trace-m1/retained/public-report.json), [Provenance](interop/trace-m1/retained/QUALIFICATION.json) and [Contract](https://github.com/darklordVirtual/REMORA-research/pull/822). [Recorded procedure](trace-m1-consumer.md) ### Per-claim results | Claim | Recorded result | |--------------------------------------|-----------------| | frozen-case-population | pass | | declared-comparison-agreement | pass | | native-contradiction-preserved | pass | | absent-selected-report-refused | pass | | trace-trust-record-ceiling-preserved | pass | | native-result-truth | not-exercised | | outside-host-adoption | not-exercised | | trace-conformance | out-of-scope | ### Roles | Role | Declaration | |----|----| | readerAuthorship | Probity; second implementation maintained outside REMORA | | fixtureOperator | Probity | | evidenceVantage | artifact | | producerIndependence | external implementation and operator; no REMORA runtime or reference-verifier imports | | independentWitness | not-established | ### Limits - Bounded ASCII JSON, explicit report digest, P1Y retention and frozen synthetic inputs - Native reason codes are preserved; independent refusal codes are not equated - Signature and binding checks do not establish truth, workload identity, authority or effects - Unsigned external-evidence mapping is not a TRACE Trust Record or conformance level - No REMORA or AgenTrust host adoption, deployment or commercial agreement is established answerExposure: Author read producer source and expected cases before implementation; not an answer-blind study comparisonOrder: Synthetic known-answer replay; retained native output compared to published fixture expectations
[HTML view](runs.html) | [Agent guide](llms.txt)