{
  "schema_version": "ta-amps.sira-rag-benchmark.v1",
  "mode": "local_verified",
  "inspired_by": [
    "Meta SIRA (June 2026): Corpus-Discriminative Offline Enrichment for Single-Shot Retrieval",
    "RA-RFT: Structural Problem-Isomorphism Retrieval Across Surface Domains",
    "MR-Search: Multi-Round Execution Reflection into Deterministic Routing Rules",
    "HyperAgents / DGM-H (March 2026): Bounded Self-Improvement Gated by Independent Verifier"
  ],
  "corpus_size_packs": 79,
  "query_count": 60,
  "query_breakdown": {
    "colloquial_vocabulary_bridge": 30,
    "structural_analogy": 30
  },
  "arms": {
    "baseline_lexical": {
      "recall_at_1": 0.2333,
      "recall_at_3": 0.35,
      "mrr": 0.341,
      "avg_search_rounds": 3.65,
      "avg_latency_ms": 0.0717,
      "p95_latency_ms": 0.1164,
      "max_latency_ms": 0.2232,
      "avg_tokens_per_retrieval": 2263.0
    },
    "sira_enriched": {
      "recall_at_1": 0.9667,
      "recall_at_3": 0.9667,
      "mrr": 0.9742,
      "avg_search_rounds": 1.0,
      "avg_latency_ms": 0.096,
      "p95_latency_ms": 0.1657,
      "max_latency_ms": 0.2051,
      "avg_tokens_per_retrieval": 440.0
    },
    "sira_rartf_mrsearch": {
      "recall_at_1": 1.0,
      "recall_at_3": 1.0,
      "mrr": 1.0,
      "avg_search_rounds": 1.0,
      "avg_latency_ms": 0.2883,
      "p95_latency_ms": 0.4561,
      "max_latency_ms": 0.491,
      "avg_tokens_per_retrieval": 390.0
    }
  },
  "headline_deltas": {
    "recall_at_1_gain": 0.7667,
    "recall_at_3_gain": 0.65,
    "mrr_gain": 0.659,
    "search_rounds_before": 3.65,
    "search_rounds_after": 1.0,
    "token_reduction_pct": 82.77,
    "all_queries_under_2ms": true
  },
  "herma_shim_compilation": {
    "compiled_rule_count": 10,
    "avg_shim_eval_latency_ms": 0.02493,
    "max_shim_eval_latency_ms": 0.03983,
    "within_2ms_budget": true,
    "sample_rules": [
      {
        "shim_id": "SHIM-001-atomic-step-contract",
        "target_pack_id": "atomic-step-contract",
        "structural_frame": "exclusive-lease-fencing",
        "trigger_terms": [
          "substation",
          "breaker",
          "command",
          "telemetry",
          "packet",
          "unbound",
          "exit"
        ],
        "min_term_hits": 3,
        "context_prune_roles": [
          "example"
        ],
        "keep_roles": [
          "guardrail",
          "spec",
          "rules",
          "tool_card",
          "template"
        ],
        "failure_guardrail": "Operator used substation control vocabulary rather than step-contract terms.",
        "avg_token_savings": 1980,
        "supporting_episodes": 1,
        "max_eval_budget_ms": 2.0,
        "rule_sha256": "299c4de690b0aae0bcba6dac9b609e901ee429116291b3269d4272b962937fd9"
      },
      {
        "shim_id": "SHIM-002-checks-of-checks",
        "target_pack_id": "checks-of-checks",
        "structural_frame": "adversarial-mutation-kill-check",
        "trigger_terms": [
          "municipal",
          "water",
          "synthetic",
          "poisoned",
          "poison",
          "samples",
          "sampl",
          "grading",
          "grad",
          "inspector"
        ],
        "min_term_hits": 3,
        "context_prune_roles": [
          "example"
        ],
        "keep_roles": [
          "guardrail",
          "spec",
          "rules",
          "tool_card",
          "template"
        ],
        "failure_guardrail": "Lab contamination phrasing missed meta-check title.",
        "avg_token_savings": 2055,
        "supporting_episodes": 1,
        "max_eval_budget_ms": 2.0,
        "rule_sha256": "231dc13a601023bee89fa4aa23eea6d7f47d49285f736d48d72381e9ef898093"
      },
      {
        "shim_id": "SHIM-003-continuous-tuning",
        "target_pack_id": "continuous-tuning",
        "structural_frame": "paired-bootstrap-holdout-gate",
        "trigger_terms": [
          "radiology",
          "classifier",
          "bucket",
          "promotion",
          "base",
          "integrity",
          "arguments",
          "argument"
        ],
        "min_term_hits": 3,
        "context_prune_roles": [
          "example"
        ],
        "keep_roles": [
          "guardrail",
          "spec",
          "rules",
          "tool_card",
          "template"
        ],
        "failure_guardrail": "Surface summary lacked bucket promotion vocabulary.",
        "avg_token_savings": 1480,
        "supporting_episodes": 1,
        "max_eval_budget_ms": 2.0,
        "rule_sha256": "fc7d235704f92c4b4ed0fffd76634dce26868646d994317a95a03c78e828bf7f"
      },
      {
        "shim_id": "SHIM-004-hybrid-retrieval-context",
        "target_pack_id": "hybrid-retrieval-context-knapsack",
        "structural_frame": "budgeted-knapsack-selection",
        "trigger_terms": [
          "airfield",
          "cargo",
          "medical",
          "pallets",
          "pallet",
          "ceiling",
          "ceil",
          "duplicate",
          "manifests",
          "manifest"
        ],
        "min_term_hits": 3,
        "context_prune_roles": [
          "example"
        ],
        "keep_roles": [
          "guardrail",
          "spec",
          "rules",
          "tool_card",
          "template"
        ],
        "failure_guardrail": "Logistics knapsack metaphor missed retrieval surface keywords.",
        "avg_token_savings": 2115,
        "supporting_episodes": 1,
        "max_eval_budget_ms": 2.0,
        "rule_sha256": "f2e98fa052ea9e044212387a81f538afd36e48bddb60fc255e0ca5b12224b411"
      },
      {
        "shim_id": "SHIM-005-leakage-safe-regime-back",
        "target_pack_id": "leakage-safe-regime-backtesting",
        "structural_frame": "purged-temporal-regime-audit",
        "trigger_terms": [
          "oncology",
          "longitudinal",
          "biomarker",
          "horizon",
          "causal",
          "bars",
          "drag"
        ],
        "min_term_hits": 3,
        "context_prune_roles": [
          "example"
        ],
        "keep_roles": [
          "guardrail",
          "spec",
          "rules",
          "tool_card",
          "template"
        ],
        "failure_guardrail": "Clinical biomarker terminology missed financial backtesting title.",
        "avg_token_savings": 1545,
        "supporting_episodes": 1,
        "max_eval_budget_ms": 2.0,
        "rule_sha256": "781d23b9405ec05448c3dbf668d519d746fa3ffd54d1f52bbd0733b86fcd6e12"
      },
      {
        "shim_id": "SHIM-006-partner-compute-cascade",
        "target_pack_id": "partner-compute-cascade",
        "structural_frame": "rate-paced-burst-fanout",
        "trigger_terms": [
          "cyclone",
          "alert",
          "burst",
          "parallel",
          "sliding",
          "slid",
          "window",
          "lease"
        ],
        "min_term_hits": 3,
        "context_prune_roles": [
          "example"
        ],
        "keep_roles": [
          "guardrail",
          "spec",
          "rules",
          "tool_card",
          "template"
        ],
        "failure_guardrail": "Disaster burst phrasing initially hit flyer pack instead of compute swarm.",
        "avg_token_savings": 1470,
        "supporting_episodes": 1,
        "max_eval_budget_ms": 2.0,
        "rule_sha256": "db9823dbd67d634bffad7d09057986eef0d7ab22a4a411ae2e96a04d78d6fc4e"
      },
      {
        "shim_id": "SHIM-007-persona-panel-second-opi",
        "target_pack_id": "persona-panel-second-opinion",
        "structural_frame": "independent-two-step-review",
        "trigger_terms": [
          "compounding",
          "compound",
          "pharmacy",
          "separate",
          "shifts",
          "shift",
          "sign",
          "syringe"
        ],
        "min_term_hits": 3,
        "context_prune_roles": [
          "example"
        ],
        "keep_roles": [
          "guardrail",
          "spec",
          "rules",
          "tool_card",
          "template"
        ],
        "failure_guardrail": "Surface lexical search missed pharmacy separation-of-duties phrasing for 3 rounds.",
        "avg_token_savings": 2090,
        "supporting_episodes": 1,
        "max_eval_budget_ms": 2.0,
        "rule_sha256": "61c2da94481fcba3e4c68b48792135cc897514a6bf6b379c52ae347a71e0eeb0"
      },
      {
        "shim_id": "SHIM-008-public-good-kits",
        "target_pack_id": "public-good-kits",
        "structural_frame": "split-sensitive-vs-public",
        "trigger_terms": [
          "maternal",
          "health",
          "clinic",
          "confidential",
          "notebooks",
          "notebook",
          "laptop",
          "bilingual",
          "criteria"
        ],
        "min_term_hits": 3,
        "context_prune_roles": [
          "example"
        ],
        "keep_roles": [
          "guardrail",
          "spec",
          "rules",
          "tool_card",
          "template"
        ],
        "failure_guardrail": "Medical clinic phrasing required 4 rounds to locate public-good kit boundary.",
        "avg_token_savings": 2090,
        "supporting_episodes": 1,
        "max_eval_budget_ms": 2.0,
        "rule_sha256": "8023180dc1b6acc06f97410865f6504818720c204f931763dd3a4260c7d76d1b"
      }
    ]
  },
  "dgm_h_verification": {
    "total_proposals": 10,
    "verified_ready_count": 10,
    "self_approval_refusal_probe": {
      "proposal_id": "DGMH-PROP-001",
      "proposer_worker_id": "worker-sira-proposer-01",
      "verifier_worker_id": "worker-sira-proposer-01",
      "distinct_actor_verified": false,
      "approved_for_owner_review": false,
      "http_status_equivalent": 409,
      "verdict": "rejected",
      "failed_checks": [
        "self_verification_forbidden_proposer_equals_verifier"
      ],
      "measured_shim_latency_ms": 0.08,
      "verification_sha256": "e1ba26ca6d79fac5e6be68a77e0d40501052f79ed3e587557fe25e40a4dddeef"
    },
    "sample_verified_proposals": [
      {
        "proposal": {
          "proposal_id": "DGMH-PROP-001",
          "framework": "HyperAgents-DGM-H-bounded",
          "proposer_worker_id": "worker-sira-proposer-01",
          "proposer_key_id": "key-sira-proposer-01",
          "target_pack_id": "atomic-step-contract",
          "patch_kind": "sira_pqe_and_shim_rule_update",
          "added_pqe_phrases": [
            "substation breaker command telemetry packet unbound exit"
          ],
          "mitigated_failure_mode": "Operator used substation control vocabulary rather than step-contract terms.",
          "evidence_episode_count": 1,
          "visibility_scope": "surface_only",
          "mutates_evaluator_controls": false,
          "holdout_mrr_delta": 0.2,
          "holdout_regressions": 0,
          "status": "quarantined_proposal",
          "proposal_sha256": "34509ec249a3ce138e83583f195e907d778b487db720d61b3c3b800adb7b0a10"
        },
        "verification": {
          "proposal_id": "DGMH-PROP-001",
          "proposer_worker_id": "worker-sira-proposer-01",
          "verifier_worker_id": "worker-dgmh-verifier-02",
          "distinct_actor_verified": true,
          "approved_for_owner_review": true,
          "http_status_equivalent": 200,
          "verdict": "verified_ready_for_owner",
          "failed_checks": [],
          "measured_shim_latency_ms": 0.06,
          "verification_sha256": "dc3619458679322869ec8a829a3529f8832143d708c8c9fc85d771b03bb74646"
        }
      },
      {
        "proposal": {
          "proposal_id": "DGMH-PROP-002",
          "framework": "HyperAgents-DGM-H-bounded",
          "proposer_worker_id": "worker-sira-proposer-01",
          "proposer_key_id": "key-sira-proposer-01",
          "target_pack_id": "checks-of-checks",
          "patch_kind": "sira_pqe_and_shim_rule_update",
          "added_pqe_phrases": [
            "municipal water lab synthetic poisoned samples grading inspector"
          ],
          "mitigated_failure_mode": "Lab contamination phrasing missed meta-check title.",
          "evidence_episode_count": 1,
          "visibility_scope": "surface_only",
          "mutates_evaluator_controls": false,
          "holdout_mrr_delta": 0.2,
          "holdout_regressions": 0,
          "status": "quarantined_proposal",
          "proposal_sha256": "1a572f6c701cdd79ff3cab2a5d92e68b6dfe1736c64f0011803e2112b33362d4"
        },
        "verification": {
          "proposal_id": "DGMH-PROP-002",
          "proposer_worker_id": "worker-sira-proposer-01",
          "verifier_worker_id": "worker-dgmh-verifier-02",
          "distinct_actor_verified": true,
          "approved_for_owner_review": true,
          "http_status_equivalent": 200,
          "verdict": "verified_ready_for_owner",
          "failed_checks": [],
          "measured_shim_latency_ms": 0.06,
          "verification_sha256": "5d31d66500d80b21f98fb5905c67ad3acc727035f36bbc429706d2e8d18ceb5a"
        }
      },
      {
        "proposal": {
          "proposal_id": "DGMH-PROP-003",
          "framework": "HyperAgents-DGM-H-bounded",
          "proposer_worker_id": "worker-sira-proposer-01",
          "proposer_key_id": "key-sira-proposer-01",
          "target_pack_id": "continuous-tuning",
          "patch_kind": "sira_pqe_and_shim_rule_update",
          "added_pqe_phrases": [
            "radiology classifier bucket promotion base integrity arguments"
          ],
          "mitigated_failure_mode": "Surface summary lacked bucket promotion vocabulary.",
          "evidence_episode_count": 1,
          "visibility_scope": "surface_only",
          "mutates_evaluator_controls": false,
          "holdout_mrr_delta": 0.2,
          "holdout_regressions": 0,
          "status": "quarantined_proposal",
          "proposal_sha256": "3090a774ccd632528ebd2392674c6dfaa451056b5b55905c121d9c701c0d4c1a"
        },
        "verification": {
          "proposal_id": "DGMH-PROP-003",
          "proposer_worker_id": "worker-sira-proposer-01",
          "verifier_worker_id": "worker-dgmh-verifier-02",
          "distinct_actor_verified": true,
          "approved_for_owner_review": true,
          "http_status_equivalent": 200,
          "verdict": "verified_ready_for_owner",
          "failed_checks": [],
          "measured_shim_latency_ms": 0.06,
          "verification_sha256": "3adb0da90e4e7e5d56e34e44e8da45d2712e526b67719e1bc1ffb49f8303b3f4"
        }
      },
      {
        "proposal": {
          "proposal_id": "DGMH-PROP-004",
          "framework": "HyperAgents-DGM-H-bounded",
          "proposer_worker_id": "worker-sira-proposer-01",
          "proposer_key_id": "key-sira-proposer-01",
          "target_pack_id": "hybrid-retrieval-context-knapsack",
          "patch_kind": "sira_pqe_and_shim_rule_update",
          "added_pqe_phrases": [
            "airfield cargo medical pallets ceiling duplicate manifests"
          ],
          "mitigated_failure_mode": "Logistics knapsack metaphor missed retrieval surface keywords.",
          "evidence_episode_count": 1,
          "visibility_scope": "surface_only",
          "mutates_evaluator_controls": false,
          "holdout_mrr_delta": 0.2,
          "holdout_regressions": 0,
          "status": "quarantined_proposal",
          "proposal_sha256": "027b29c09b86a5e6704bc2ab4d8ffc00d2e0e9ca1360e399a0ea0cccddeb36ac"
        },
        "verification": {
          "proposal_id": "DGMH-PROP-004",
          "proposer_worker_id": "worker-sira-proposer-01",
          "verifier_worker_id": "worker-dgmh-verifier-02",
          "distinct_actor_verified": true,
          "approved_for_owner_review": true,
          "http_status_equivalent": 200,
          "verdict": "verified_ready_for_owner",
          "failed_checks": [],
          "measured_shim_latency_ms": 0.06,
          "verification_sha256": "b684c762418edfc686cdeaf29524a9cb09f62fc5d018d03e86af34bbbb445655"
        }
      },
      {
        "proposal": {
          "proposal_id": "DGMH-PROP-005",
          "framework": "HyperAgents-DGM-H-bounded",
          "proposer_worker_id": "worker-sira-proposer-01",
          "proposer_key_id": "key-sira-proposer-01",
          "target_pack_id": "leakage-safe-regime-backtesting",
          "patch_kind": "sira_pqe_and_shim_rule_update",
          "added_pqe_phrases": [
            "oncology longitudinal biomarker horizon causal bars drag"
          ],
          "mitigated_failure_mode": "Clinical biomarker terminology missed financial backtesting title.",
          "evidence_episode_count": 1,
          "visibility_scope": "surface_only",
          "mutates_evaluator_controls": false,
          "holdout_mrr_delta": 0.2,
          "holdout_regressions": 0,
          "status": "quarantined_proposal",
          "proposal_sha256": "9ac3eea1e94d9857321571444439a5ae7dedcd8301f06709700b166ca5bf8b57"
        },
        "verification": {
          "proposal_id": "DGMH-PROP-005",
          "proposer_worker_id": "worker-sira-proposer-01",
          "verifier_worker_id": "worker-dgmh-verifier-02",
          "distinct_actor_verified": true,
          "approved_for_owner_review": true,
          "http_status_equivalent": 200,
          "verdict": "verified_ready_for_owner",
          "failed_checks": [],
          "measured_shim_latency_ms": 0.06,
          "verification_sha256": "a96c1fb074e4d4beac9377f91c5e6648a436ef7c5d9e0088ffb0071734549202"
        }
      },
      {
        "proposal": {
          "proposal_id": "DGMH-PROP-006",
          "framework": "HyperAgents-DGM-H-bounded",
          "proposer_worker_id": "worker-sira-proposer-01",
          "proposer_key_id": "key-sira-proposer-01",
          "target_pack_id": "partner-compute-cascade",
          "patch_kind": "sira_pqe_and_shim_rule_update",
          "added_pqe_phrases": [
            "cyclone alert burst parallel workers sliding window lease"
          ],
          "mitigated_failure_mode": "Disaster burst phrasing initially hit flyer pack instead of compute swarm.",
          "evidence_episode_count": 1,
          "visibility_scope": "surface_only",
          "mutates_evaluator_controls": false,
          "holdout_mrr_delta": 0.2,
          "holdout_regressions": 0,
          "status": "quarantined_proposal",
          "proposal_sha256": "340d4307624226d1227b45a717a3bdb940ea79ba90d80c71c216ae85377903a6"
        },
        "verification": {
          "proposal_id": "DGMH-PROP-006",
          "proposer_worker_id": "worker-sira-proposer-01",
          "verifier_worker_id": "worker-dgmh-verifier-02",
          "distinct_actor_verified": true,
          "approved_for_owner_review": true,
          "http_status_equivalent": 200,
          "verdict": "verified_ready_for_owner",
          "failed_checks": [],
          "measured_shim_latency_ms": 0.06,
          "verification_sha256": "ac4fa15c2f35279a9c2d73b6fb2a1840c56487658d0cd194e0d89daa648494ef"
        }
      }
    ]
  },
  "per_query_results": [
    {
      "id": "Q01",
      "kind": "colloquial_vocabulary_bridge",
      "query": "Smudged handwritten intake notes with crossed-out abbreviations and illegible lines",
      "relevant_packs": [
        "kit-handwritten-statements"
      ],
      "baseline_top1": "exploratory-testing-charters",
      "baseline_rank": 3,
      "baseline_rounds": 3,
      "sira_top1": "kit-handwritten-statements",
      "sira_rank": 1,
      "sira_rartf_top1": "kit-handwritten-statements",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.4144
    },
    {
      "id": "Q02",
      "kind": "colloquial_vocabulary_bridge",
      "query": "Dark theme foreground footer a11y conformance and assistive selector width",
      "relevant_packs": [
        "accessibility-audit"
      ],
      "baseline_top1": "wiring-and-conformance-checks",
      "baseline_rank": 4,
      "baseline_rounds": 4,
      "sira_top1": "accessibility-audit",
      "sira_rank": 1,
      "sira_rartf_top1": "accessibility-audit",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.1997
    },
    {
      "id": "Q03",
      "kind": "colloquial_vocabulary_bridge",
      "query": "Background music ducking envelope, integrated LUFS metrics and pause rise",
      "relevant_packs": [
        "audio-and-narration-qa"
      ],
      "baseline_top1": "video-editing",
      "baseline_rank": 4,
      "baseline_rounds": 4,
      "sira_top1": "audio-and-narration-qa",
      "sira_rank": 1,
      "sira_rartf_top1": "audio-and-narration-qa",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.1971
    },
    {
      "id": "Q04",
      "kind": "colloquial_vocabulary_bridge",
      "query": "Ken Burns slide pan drift, static terminal glass pill and shot dwell journeys",
      "relevant_packs": [
        "scene-and-cognitive-video-qa"
      ],
      "baseline_top1": "codebase-health-review",
      "baseline_rank": 2,
      "baseline_rounds": 3,
      "sira_top1": "scene-and-cognitive-video-qa",
      "sira_rank": 1,
      "sira_rartf_top1": "scene-and-cognitive-video-qa",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.3013
    },
    {
      "id": "Q05",
      "kind": "colloquial_vocabulary_bridge",
      "query": "Sanitized embedding chunks with redundancy alarm and injected directives redact",
      "relevant_packs": [
        "hybrid-retrieval-context-knapsack"
      ],
      "baseline_top1": "accessibility-audit",
      "baseline_rank": 28,
      "baseline_rounds": 5,
      "sira_top1": "hybrid-retrieval-context-knapsack",
      "sira_rank": 1,
      "sira_rartf_top1": "hybrid-retrieval-context-knapsack",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.2641
    },
    {
      "id": "Q06",
      "kind": "colloquial_vocabulary_bridge",
      "query": "Ngram Jaccard disjointness against holdouts with supersession churn and abstention",
      "relevant_packs": [
        "lora-memory-and-shimming"
      ],
      "baseline_top1": "lora-memory-and-shimming",
      "baseline_rank": 1,
      "baseline_rounds": 1,
      "sira_top1": "lora-memory-and-shimming",
      "sira_rank": 1,
      "sira_rartf_top1": "lora-memory-and-shimming",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.2962
    },
    {
      "id": "Q07",
      "kind": "colloquial_vocabulary_bridge",
      "query": "Realized Sharpe causal horizon bars with deflated prediction series drag",
      "relevant_packs": [
        "leakage-safe-regime-backtesting"
      ],
      "baseline_top1": "accessibility-audit",
      "baseline_rank": 39,
      "baseline_rounds": 5,
      "sira_top1": "leakage-safe-regime-backtesting",
      "sira_rank": 1,
      "sira_rartf_top1": "leakage-safe-regime-backtesting",
      "sira_rartf_rank": 1,
      "matched_shims": [
        "SHIM-005-leakage-safe-regime-back"
      ],
      "sira_rartf_latency_ms": 0.1853
    },
    {
      "id": "Q08",
      "kind": "colloquial_vocabulary_bridge",
      "query": "Kill operators across targeted validation topics for trusted release checks",
      "relevant_packs": [
        "checks-of-checks"
      ],
      "baseline_top1": "release-and-verify",
      "baseline_rank": 2,
      "baseline_rounds": 3,
      "sira_top1": "checks-of-checks",
      "sira_rank": 1,
      "sira_rartf_top1": "checks-of-checks",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.329
    },
    {
      "id": "Q09",
      "kind": "colloquial_vocabulary_bridge",
      "query": "Upstream unreachable watched notices with staleness schedule and replacement actions",
      "relevant_packs": [
        "freshness-checks"
      ],
      "baseline_top1": "freshness-checks",
      "baseline_rank": 1,
      "baseline_rounds": 1,
      "sira_top1": "freshness-checks",
      "sira_rank": 1,
      "sira_rartf_top1": "freshness-checks",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.184
    },
    {
      "id": "Q10",
      "kind": "colloquial_vocabulary_bridge",
      "query": "Adopted FY2026 FY2027 fiscal literacy budget figures across municipal services",
      "relevant_packs": [
        "public-budget-report"
      ],
      "baseline_top1": "kit-funding-impact-summary",
      "baseline_rank": 4,
      "baseline_rounds": 4,
      "sira_top1": "public-budget-report",
      "sira_rank": 1,
      "sira_rartf_top1": "public-budget-report",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.1785
    },
    {
      "id": "Q11",
      "kind": "colloquial_vocabulary_bridge",
      "query": "Multilingual clinic justice postings with language checker corrections and questions",
      "relevant_packs": [
        "kit-flyer-translation-check"
      ],
      "baseline_top1": "repo-review",
      "baseline_rank": 8,
      "baseline_rounds": 4,
      "sira_top1": "kit-flyer-translation-check",
      "sira_rank": 1,
      "sira_rartf_top1": "kit-flyer-translation-check",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.3595
    },
    {
      "id": "Q12",
      "kind": "colloquial_vocabulary_bridge",
      "query": "Weekly pantry cooler sheet moves with acceptable scheduling windows and codes",
      "relevant_packs": [
        "kit-food-bank-pickups"
      ],
      "baseline_top1": "file-triage-at-scale",
      "baseline_rank": 4,
      "baseline_rounds": 4,
      "sira_top1": "kit-food-bank-pickups",
      "sira_rank": 1,
      "sira_rartf_top1": "kit-food-bank-pickups",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.1816
    },
    {
      "id": "Q13",
      "kind": "colloquial_vocabulary_bridge",
      "query": "Anonymous mail merge greeting rows with cleared goods vs services development choice",
      "relevant_packs": [
        "kit-donor-thank-you-letters"
      ],
      "baseline_top1": "kit-victim-support-referrals",
      "baseline_rank": 33,
      "baseline_rounds": 5,
      "sira_top1": "kit-donor-thank-you-letters",
      "sira_rank": 1,
      "sira_rartf_top1": "kit-donor-thank-you-letters",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.1946
    },
    {
      "id": "Q14",
      "kind": "colloquial_vocabulary_bridge",
      "query": "Unmatched adult gear assignments and matcher brief suggestions for relief meeting",
      "relevant_packs": [
        "kit-volunteer-shift-schedule"
      ],
      "baseline_top1": "kit-policy-research-brief",
      "baseline_rank": 2,
      "baseline_rounds": 3,
      "sira_top1": "kit-volunteer-shift-schedule",
      "sira_rank": 1,
      "sira_rartf_top1": "kit-volunteer-shift-schedule",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.1734
    },
    {
      "id": "Q15",
      "kind": "colloquial_vocabulary_bridge",
      "query": "Advocate counseling consent planning story details in survivor preferred languages",
      "relevant_packs": [
        "kit-victim-support-referrals"
      ],
      "baseline_top1": "parametric-3d-video-composer",
      "baseline_rank": 38,
      "baseline_rounds": 5,
      "sira_top1": "kit-victim-support-referrals",
      "sira_rank": 1,
      "sira_rartf_top1": "kit-victim-support-referrals",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.1745
    },
    {
      "id": "Q16",
      "kind": "colloquial_vocabulary_bridge",
      "query": "FTC advance-fee loan deposit red flags, coercive payment demand and payee jurisdiction",
      "relevant_packs": [
        "exploitation-evidence"
      ],
      "baseline_top1": "benchmark-scores",
      "baseline_rank": 22,
      "baseline_rounds": 5,
      "sira_top1": "exploitation-evidence",
      "sira_rank": 1,
      "sira_rartf_top1": "exploitation-evidence",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.1811
    },
    {
      "id": "Q17",
      "kind": "colloquial_vocabulary_bridge",
      "query": "92pct GPU VRAM Glimmer fit, container isolation flight gates and statute citation rubric",
      "relevant_packs": [
        "partner-compute-cascade"
      ],
      "baseline_top1": "lora-memory-and-shimming",
      "baseline_rank": 53,
      "baseline_rounds": 5,
      "sira_top1": "partner-compute-cascade",
      "sira_rank": 1,
      "sira_rartf_top1": "partner-compute-cascade",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.3067
    },
    {
      "id": "Q18",
      "kind": "colloquial_vocabulary_bridge",
      "query": "Airgap unsplit cloud egress block across four-axis episode sweep and candidate digestion",
      "relevant_packs": [
        "meta-are-scenario-matrix"
      ],
      "baseline_top1": "rename-and-relink",
      "baseline_rank": 5,
      "baseline_rounds": 4,
      "sira_top1": "meta-are-scenario-matrix",
      "sira_rank": 1,
      "sira_rartf_top1": "meta-are-scenario-matrix",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.1919
    },
    {
      "id": "Q19",
      "kind": "colloquial_vocabulary_bridge",
      "query": "Predictive query expansions, top1 rartf analogy lift and evaluator holdout protection",
      "relevant_packs": [
        "sira-self-learning-rag"
      ],
      "baseline_top1": "usability-heuristics-review",
      "baseline_rank": 63,
      "baseline_rounds": 5,
      "sira_top1": "sira-self-learning-rag",
      "sira_rank": 1,
      "sira_rartf_top1": "sira-self-learning-rag",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.4044
    },
    {
      "id": "Q20",
      "kind": "colloquial_vocabulary_bridge",
      "query": "Pending broad handler suppress markers and authored roster sync locations",
      "relevant_packs": [
        "codebase-health-review"
      ],
      "baseline_top1": "screenshot-design-review",
      "baseline_rank": 11,
      "baseline_rounds": 5,
      "sira_top1": "codebase-health-review",
      "sira_rank": 1,
      "sira_rartf_top1": "codebase-health-review",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.1829
    },
    {
      "id": "Q21",
      "kind": "colloquial_vocabulary_bridge",
      "query": "Shallow readyz trip on dark pantry module with mocked dependency patches stopped",
      "relevant_packs": [
        "wiring-and-conformance-checks"
      ],
      "baseline_top1": "wiring-and-conformance-checks",
      "baseline_rank": 1,
      "baseline_rounds": 1,
      "sira_top1": "wiring-and-conformance-checks",
      "sira_rank": 1,
      "sira_rartf_top1": "wiring-and-conformance-checks",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.1721
    },
    {
      "id": "Q22",
      "kind": "colloquial_vocabulary_bridge",
      "query": "Pending user complaints heuristic debug mission roster and printed session notes",
      "relevant_packs": [
        "annoyance-hunt"
      ],
      "baseline_top1": "exploratory-testing-charters",
      "baseline_rank": 11,
      "baseline_rounds": 5,
      "sira_top1": "annoyance-hunt",
      "sira_rank": 1,
      "sira_rartf_top1": "annoyance-hunt",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.1722
    },
    {
      "id": "Q23",
      "kind": "colloquial_vocabulary_bridge",
      "query": "Nightly green rung integrity log with declined void and substitution rounding kept",
      "relevant_packs": [
        "unattended-run-record"
      ],
      "baseline_top1": "continuous-tuning",
      "baseline_rank": 3,
      "baseline_rounds": 3,
      "sira_top1": "unattended-run-record",
      "sira_rank": 1,
      "sira_rartf_top1": "unattended-run-record",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.1779
    },
    {
      "id": "Q24",
      "kind": "colloquial_vocabulary_bridge",
      "query": "Messy shared filesystem folder archive campaign with zero deletions",
      "relevant_packs": [
        "file-triage-at-scale"
      ],
      "baseline_top1": "kit-policy-research-brief",
      "baseline_rank": 27,
      "baseline_rounds": 5,
      "sira_top1": "file-triage-at-scale",
      "sira_rank": 1,
      "sira_rartf_top1": "file-triage-at-scale",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.3347
    },
    {
      "id": "Q25",
      "kind": "colloquial_vocabulary_bridge",
      "query": "FFmpeg lavfi freezedetect and astats codec stream warning rows",
      "relevant_packs": [
        "video-qa",
        "frame-by-frame-video-qa"
      ],
      "baseline_top1": "accessibility-audit",
      "baseline_rank": 24,
      "baseline_rounds": 5,
      "sira_top1": "video-qa",
      "sira_rank": 1,
      "sira_rartf_top1": "video-qa",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.2763
    },
    {
      "id": "Q26",
      "kind": "colloquial_vocabulary_bridge",
      "query": "IPA phoneme respelling dictionary and audition take recognizer at target WPM",
      "relevant_packs": [
        "voice-narration"
      ],
      "baseline_top1": "audio-and-narration-qa",
      "baseline_rank": 78,
      "baseline_rounds": 5,
      "sira_top1": "voice-narration",
      "sira_rank": 1,
      "sira_rartf_top1": "voice-narration",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.2649
    },
    {
      "id": "Q27",
      "kind": "colloquial_vocabulary_bridge",
      "query": "Wilson intervals normalizing frontier coder benchmarks across child matrix tasks",
      "relevant_packs": [
        "agent-harness-feeds"
      ],
      "baseline_top1": "meta-are-scenario-matrix",
      "baseline_rank": 4,
      "baseline_rounds": 4,
      "sira_top1": "agent-harness-feeds",
      "sira_rank": 1,
      "sira_rartf_top1": "agent-harness-feeds",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.2029
    },
    {
      "id": "Q28",
      "kind": "colloquial_vocabulary_bridge",
      "query": "Vendor leaderboard paper subjects flagging contradicted or unverified reporter fetches",
      "relevant_packs": [
        "benchmark-scores"
      ],
      "baseline_top1": "kit-handwritten-statements",
      "baseline_rank": 11,
      "baseline_rounds": 5,
      "sira_top1": "benchmark-scores",
      "sira_rank": 1,
      "sira_rartf_top1": "benchmark-scores",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.1689
    },
    {
      "id": "Q29",
      "kind": "colloquial_vocabulary_bridge",
      "query": "Cloud cost invoices currency pricing quantity basis and period trade hypotheses",
      "relevant_packs": [
        "cost-optimization"
      ],
      "baseline_top1": "cost-optimization",
      "baseline_rank": 1,
      "baseline_rounds": 1,
      "sira_top1": "cost-optimization",
      "sira_rank": 1,
      "sira_rartf_top1": "cost-optimization",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.1664
    },
    {
      "id": "Q30",
      "kind": "colloquial_vocabulary_bridge",
      "query": "City council food distribution ordinances tracking mandatory shall permit registration",
      "relevant_packs": [
        "policy-and-law-watch",
        "kit-policy-research-brief"
      ],
      "baseline_top1": "kit-food-bank-pickups",
      "baseline_rank": 38,
      "baseline_rounds": 5,
      "sira_top1": "policy-and-law-watch",
      "sira_rank": 1,
      "sira_rartf_top1": "policy-and-law-watch",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.2002
    },
    {
      "id": "Q31",
      "kind": "structural_analogy",
      "query": "Maternal health clinic keeps confidential notebooks on its local laptop while bilingual volunteers translate manifest criteria",
      "relevant_packs": [
        "public-good-kits",
        "kit-handwritten-statements",
        "kit-victim-support-referrals"
      ],
      "baseline_top1": "kit-volunteer-shift-schedule",
      "baseline_rank": 4,
      "baseline_rounds": 4,
      "sira_top1": "public-good-kits",
      "sira_rank": 1,
      "sira_rartf_top1": "public-good-kits",
      "sira_rartf_rank": 1,
      "matched_shims": [
        "SHIM-008-public-good-kits"
      ],
      "sira_rartf_latency_ms": 0.4168
    },
    {
      "id": "Q32",
      "kind": "structural_analogy",
      "query": "Substation breaker command requires an assembled telemetry packet and unbound runner exit",
      "relevant_packs": [
        "atomic-step-contract"
      ],
      "baseline_top1": "parametric-3d-video-composer",
      "baseline_rank": 7,
      "baseline_rounds": 4,
      "sira_top1": "atomic-step-contract",
      "sira_rank": 1,
      "sira_rartf_top1": "atomic-step-contract",
      "sira_rartf_rank": 1,
      "matched_shims": [
        "SHIM-001-atomic-step-contract"
      ],
      "sira_rartf_latency_ms": 0.2992
    },
    {
      "id": "Q33",
      "kind": "structural_analogy",
      "query": "Compounding pharmacy requires separate shifts to sign off a syringe mix with skeptic factual positions",
      "relevant_packs": [
        "persona-panel-second-opinion"
      ],
      "baseline_top1": "kit-food-bank-pickups",
      "baseline_rank": 49,
      "baseline_rounds": 5,
      "sira_top1": "persona-panel-second-opinion",
      "sira_rank": 1,
      "sira_rartf_top1": "persona-panel-second-opinion",
      "sira_rartf_rank": 1,
      "matched_shims": [
        "SHIM-007-persona-panel-second-opi"
      ],
      "sira_rartf_latency_ms": 0.4561
    },
    {
      "id": "Q34",
      "kind": "structural_analogy",
      "query": "Avionics firmware cutover switch checks landed commit sha parity conclusion",
      "relevant_packs": [
        "release-and-verify",
        "launch-readiness"
      ],
      "baseline_top1": "rename-and-relink",
      "baseline_rank": 9,
      "baseline_rounds": 5,
      "sira_top1": "release-and-verify",
      "sira_rank": 1,
      "sira_rartf_top1": "release-and-verify",
      "sira_rartf_rank": 1,
      "matched_shims": [
        "SHIM-009-release-and-verify"
      ],
      "sira_rartf_latency_ms": 0.3284
    },
    {
      "id": "Q35",
      "kind": "structural_analogy",
      "query": "Agricultural irrigation controller runs a qualification rung lookup to replay solved deliverables",
      "relevant_packs": [
        "retrieval-triggers-and-reuse",
        "prior-art-and-alternatives"
      ],
      "baseline_top1": "evaluation-design",
      "baseline_rank": 53,
      "baseline_rounds": 5,
      "sira_top1": "retrieval-triggers-and-reuse",
      "sira_rank": 1,
      "sira_rartf_top1": "retrieval-triggers-and-reuse",
      "sira_rartf_rank": 1,
      "matched_shims": [
        "SHIM-010-retrieval-triggers-and-r"
      ],
      "sira_rartf_latency_ms": 0.3761
    },
    {
      "id": "Q36",
      "kind": "structural_analogy",
      "query": "Radiology classifier bucket promotion verifies base integrity feeds and offered arguments",
      "relevant_packs": [
        "continuous-tuning",
        "lora-memory-and-shimming"
      ],
      "baseline_top1": "agent-harness-feeds",
      "baseline_rank": 2,
      "baseline_rounds": 3,
      "sira_top1": "continuous-tuning",
      "sira_rank": 1,
      "sira_rartf_top1": "continuous-tuning",
      "sira_rartf_rank": 1,
      "matched_shims": [
        "SHIM-003-continuous-tuning"
      ],
      "sira_rartf_latency_ms": 0.3068
    },
    {
      "id": "Q37",
      "kind": "structural_analogy",
      "query": "Airfield cargo dispatcher packs medical pallets under a ceiling while dropping duplicate manifests",
      "relevant_packs": [
        "hybrid-retrieval-context-knapsack"
      ],
      "baseline_top1": "evaluation-design",
      "baseline_rank": 31,
      "baseline_rounds": 5,
      "sira_top1": "evaluation-design",
      "sira_rank": 5,
      "sira_rartf_top1": "hybrid-retrieval-context-knapsack",
      "sira_rartf_rank": 1,
      "matched_shims": [
        "SHIM-004-hybrid-retrieval-context"
      ],
      "sira_rartf_latency_ms": 0.2556
    },
    {
      "id": "Q38",
      "kind": "structural_analogy",
      "query": "Oncology longitudinal biomarker cohort study checks causal horizon bars and Sharpe drag",
      "relevant_packs": [
        "leakage-safe-regime-backtesting"
      ],
      "baseline_top1": "checks-of-checks",
      "baseline_rank": 59,
      "baseline_rounds": 5,
      "sira_top1": "leakage-safe-regime-backtesting",
      "sira_rank": 1,
      "sira_rartf_top1": "leakage-safe-regime-backtesting",
      "sira_rartf_rank": 1,
      "matched_shims": [
        "SHIM-005-leakage-safe-regime-back"
      ],
      "sira_rartf_latency_ms": 0.292
    },
    {
      "id": "Q39",
      "kind": "structural_analogy",
      "query": "Vertical 9x16 transit announcement reels checking cubic easing phase curves and installation norms",
      "relevant_packs": [
        "video-motion",
        "scene-and-cognitive-video-qa"
      ],
      "baseline_top1": "parametric-3d-video-composer",
      "baseline_rank": 6,
      "baseline_rounds": 4,
      "sira_top1": "video-motion",
      "sira_rank": 1,
      "sira_rartf_top1": "video-motion",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.2905
    },
    {
      "id": "Q40",
      "kind": "structural_analogy",
      "query": "Municipal water lab synthetic poisoned samples graded by inspector kill operators",
      "relevant_packs": [
        "checks-of-checks"
      ],
      "baseline_top1": "meta-are-scenario-matrix",
      "baseline_rank": 15,
      "baseline_rounds": 5,
      "sira_top1": "checks-of-checks",
      "sira_rank": 1,
      "sira_rartf_top1": "checks-of-checks",
      "sira_rartf_rank": 1,
      "matched_shims": [
        "SHIM-002-checks-of-checks"
      ],
      "sira_rartf_latency_ms": 0.2717
    },
    {
      "id": "Q41",
      "kind": "structural_analogy",
      "query": "Cyclone alert burst dispatching async parallel workers under a sliding window lease",
      "relevant_packs": [
        "partner-compute-cascade"
      ],
      "baseline_top1": "leakage-safe-regime-backtesting",
      "baseline_rank": 47,
      "baseline_rounds": 5,
      "sira_top1": "partner-compute-cascade",
      "sira_rank": 1,
      "sira_rartf_top1": "partner-compute-cascade",
      "sira_rartf_rank": 1,
      "matched_shims": [
        "SHIM-006-partner-compute-cascade"
      ],
      "sira_rartf_latency_ms": 0.3096
    },
    {
      "id": "Q42",
      "kind": "structural_analogy",
      "query": "Customs border intake checking Unicode NFC charset strings, leap-day classes and oracle outputs",
      "relevant_packs": [
        "edge-and-corner-cases"
      ],
      "baseline_top1": "visual-rendering",
      "baseline_rank": 49,
      "baseline_rounds": 5,
      "sira_top1": "edge-and-corner-cases",
      "sira_rank": 1,
      "sira_rartf_top1": "edge-and-corner-cases",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.2954
    },
    {
      "id": "Q43",
      "kind": "structural_analogy",
      "query": "Municipal transit inspection diffing reopened defect fingerprints and overdue churn across runs",
      "relevant_packs": [
        "continuous-self-review",
        "findings-to-fixes"
      ],
      "baseline_top1": "continuous-self-review",
      "baseline_rank": 1,
      "baseline_rounds": 1,
      "sira_top1": "continuous-self-review",
      "sira_rank": 1,
      "sira_rartf_top1": "continuous-self-review",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.491
    },
    {
      "id": "Q44",
      "kind": "structural_analogy",
      "query": "Sensor firmware p75 LCP latency check within comparable bands to separate flake shifts",
      "relevant_packs": [
        "regression-watch"
      ],
      "baseline_top1": "kit-food-bank-pickups",
      "baseline_rank": 23,
      "baseline_rounds": 5,
      "sira_top1": "regression-watch",
      "sira_rank": 1,
      "sira_rartf_top1": "regression-watch",
      "sira_rartf_rank": 1,
      "matched_shims": [
        "SHIM-007-persona-panel-second-opi"
      ],
      "sira_rartf_latency_ms": 0.4019
    },
    {
      "id": "Q45",
      "kind": "structural_analogy",
      "query": "Regional public health synthesis excluding retracted papers so every sentence cites a verified study",
      "relevant_packs": [
        "deep-research-report",
        "measured-claims"
      ],
      "baseline_top1": "deep-research-report",
      "baseline_rank": 1,
      "baseline_rounds": 1,
      "sira_top1": "deep-research-report",
      "sira_rank": 1,
      "sira_rartf_top1": "deep-research-report",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.3149
    },
    {
      "id": "Q46",
      "kind": "structural_analogy",
      "query": "Epidemiology meal-pound site dataset checking column weeks across northern districts",
      "relevant_packs": [
        "checked-data-analysis"
      ],
      "baseline_top1": "visual-rendering",
      "baseline_rank": 16,
      "baseline_rounds": 5,
      "sira_top1": "checked-data-analysis",
      "sira_rank": 1,
      "sira_rartf_top1": "checked-data-analysis",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.2818
    },
    {
      "id": "Q47",
      "kind": "structural_analogy",
      "query": "Hospital board contexts and funder emails turned into versioned direction while keeping superseded guidance",
      "relevant_packs": [
        "direction-to-work"
      ],
      "baseline_top1": "direction-to-work",
      "baseline_rank": 1,
      "baseline_rounds": 1,
      "sira_top1": "direction-to-work",
      "sira_rank": 1,
      "sira_rartf_top1": "direction-to-work",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.3635
    },
    {
      "id": "Q48",
      "kind": "structural_analogy",
      "query": "Nightly operations shift-change bulletin recording observed sitemap backup pins",
      "relevant_packs": [
        "status-and-handoff"
      ],
      "baseline_top1": "user-acceptance-testing",
      "baseline_rank": 70,
      "baseline_rounds": 5,
      "sira_top1": "status-and-handoff",
      "sira_rank": 1,
      "sira_rartf_top1": "status-and-handoff",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.3189
    },
    {
      "id": "Q49",
      "kind": "structural_analogy",
      "query": "Sunrise-to-Acme repository handbook changelog variants with zero tracker violations",
      "relevant_packs": [
        "rename-and-relink"
      ],
      "baseline_top1": "repo-review",
      "baseline_rank": 57,
      "baseline_rounds": 5,
      "sira_top1": "rename-and-relink",
      "sira_rank": 1,
      "sira_rartf_top1": "rename-and-relink",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.3368
    },
    {
      "id": "Q50",
      "kind": "structural_analogy",
      "query": "Connect prewritten dedupe, impute and numeric stats nodes from src to dst ports without new code",
      "relevant_packs": [
        "code-composition"
      ],
      "baseline_top1": "code-composition",
      "baseline_rank": 1,
      "baseline_rounds": 1,
      "sira_top1": "code-composition",
      "sira_rank": 1,
      "sira_rartf_top1": "code-composition",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.3391
    },
    {
      "id": "Q51",
      "kind": "structural_analogy",
      "query": "Civic flood-barrier comparison sheet scoring airy vs formal taste preference vectors side by side",
      "relevant_packs": [
        "taste-palettes"
      ],
      "baseline_top1": "taste-palettes",
      "baseline_rank": 1,
      "baseline_rounds": 1,
      "sira_top1": "taste-palettes",
      "sira_rank": 1,
      "sira_rartf_top1": "taste-palettes",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.3705
    },
    {
      "id": "Q52",
      "kind": "structural_analogy",
      "query": "Pre-register an immutable evaluation scorer diagnosis with twin opaque controls before candidate runs",
      "relevant_packs": [
        "evaluation-design"
      ],
      "baseline_top1": "evaluation-design",
      "baseline_rank": 1,
      "baseline_rounds": 1,
      "sira_top1": "evaluation-design",
      "sira_rank": 1,
      "sira_rartf_top1": "evaluation-design",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.3311
    },
    {
      "id": "Q53",
      "kind": "structural_analogy",
      "query": "Survey existing open-source tools and non-code alternatives before building a custom municipal app",
      "relevant_packs": [
        "prior-art-and-alternatives",
        "tool-discovery"
      ],
      "baseline_top1": "prior-art-and-alternatives",
      "baseline_rank": 1,
      "baseline_rounds": 1,
      "sira_top1": "prior-art-and-alternatives",
      "sira_rank": 1,
      "sira_rartf_top1": "prior-art-and-alternatives",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.3486
    },
    {
      "id": "Q54",
      "kind": "structural_analogy",
      "query": "Time-boxed field exploratory testing session tracking overrun pct on an offline utility app",
      "relevant_packs": [
        "exploratory-testing-charters",
        "user-acceptance-testing"
      ],
      "baseline_top1": "exploratory-testing-charters",
      "baseline_rank": 1,
      "baseline_rounds": 1,
      "sira_top1": "exploratory-testing-charters",
      "sira_rank": 1,
      "sira_rartf_top1": "exploratory-testing-charters",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.3749
    },
    {
      "id": "Q55",
      "kind": "structural_analogy",
      "query": "CIK accession filer profile extracting GAAP fiscal quarter concepts and dividend factors",
      "relevant_packs": [
        "public-company-research"
      ],
      "baseline_top1": "public-budget-report",
      "baseline_rank": 5,
      "baseline_rounds": 4,
      "sira_top1": "public-company-research",
      "sira_rank": 1,
      "sira_rartf_top1": "public-company-research",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.2646
    },
    {
      "id": "Q56",
      "kind": "structural_analogy",
      "query": "Local authorized security review auditing caller authorization severity while keeping secret locations masked",
      "relevant_packs": [
        "security-review"
      ],
      "baseline_top1": "security-review",
      "baseline_rank": 1,
      "baseline_rounds": 1,
      "sira_top1": "security-review",
      "sira_rank": 1,
      "sira_rartf_top1": "security-review",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.4613
    },
    {
      "id": "Q57",
      "kind": "structural_analogy",
      "query": "Warehouse robotics cell needs single-claimant lease lock and stale command fencing on failover",
      "relevant_packs": [
        "atomic-step-contract",
        "partner-compute-cascade",
        "harness-onboarding-e2e"
      ],
      "baseline_top1": "codebase-health-review",
      "baseline_rank": 3,
      "baseline_rounds": 3,
      "sira_top1": "partner-compute-cascade",
      "sira_rank": 1,
      "sira_rartf_top1": "partner-compute-cascade",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.3503
    },
    {
      "id": "Q58",
      "kind": "structural_analogy",
      "query": "Clinical ethics board requires separation of duties where proposer worker never verifies own dgmh patch",
      "relevant_packs": [
        "sira-self-learning-rag",
        "persona-panel-second-opinion",
        "public-good-kits"
      ],
      "baseline_top1": "atomic-step-contract",
      "baseline_rank": 18,
      "baseline_rounds": 5,
      "sira_top1": "sira-self-learning-rag",
      "sira_rank": 1,
      "sira_rartf_top1": "sira-self-learning-rag",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.3611
    },
    {
      "id": "Q59",
      "kind": "structural_analogy",
      "query": "Disaster relief truck packs emergency crates under weight ceiling without redundant cargo",
      "relevant_packs": [
        "hybrid-retrieval-context-knapsack"
      ],
      "baseline_top1": "kit-volunteer-shift-schedule",
      "baseline_rank": 30,
      "baseline_rounds": 5,
      "sira_top1": "kit-volunteer-shift-schedule",
      "sira_rank": 4,
      "sira_rartf_top1": "hybrid-retrieval-context-knapsack",
      "sira_rartf_rank": 1,
      "matched_shims": [
        "SHIM-004-hybrid-retrieval-context"
      ],
      "sira_rartf_latency_ms": 0.3939
    },
    {
      "id": "Q60",
      "kind": "structural_analogy",
      "query": "River flood sensor forecast evaluated on temporal windows with lookahead leakage prevention",
      "relevant_packs": [
        "leakage-safe-regime-backtesting"
      ],
      "baseline_top1": "leakage-safe-regime-backtesting",
      "baseline_rank": 1,
      "baseline_rounds": 1,
      "sira_top1": "leakage-safe-regime-backtesting",
      "sira_rank": 1,
      "sira_rartf_top1": "leakage-safe-regime-backtesting",
      "sira_rartf_rank": 1,
      "matched_shims": [],
      "sira_rartf_latency_ms": 0.2573
    }
  ],
  "artifact_sha256": "ab678ff6542a44272e3fdbf3a0428603e1d013b95b37fc3d9efdf18d5e1c8630"
}
