{
  "schema": "prompted.forked.arxiv-metadata.v1",
  "paper_id": "P06",
  "title": "The Golden Hall Test: Meaningful Agency Under Machine-Mediated Evidence Terrain",
  "authors": [
    "Breyden E. Taylor"
  ],
  "affiliation": "Prompted LLC / Ubiquity Federation / FORKED Estate",
  "abstract": "Human presence in a workflow does not establish meaningful control. A person may approve a decision after the system selected the evidence, named the categories, ordered the alternatives, set the default, framed confidence, and made refusal operationally expensive. We define the Golden Hall Test as a reason-responsive agency evaluation over source visibility, independent formation, challenge, delay, declaration, reversal, refusal, vocabulary independence, and correction reach. The method distinguishes presence, approval, causal contribution, authority, and liability. We provide a preregistered repeated-decision study and an AgencyReceipt schema that records which rights were actually available. Synthetic pilots motivate the design but are not treated as human evidence.",
  "categories": [
    "cs.HC",
    "cs.AI",
    "cs.CY"
  ],
  "comments": "9 pages, 1 figure; preregistration, source manifest, configuration digest, and release boundary included in companion release package.",
  "date": "2026-07-24",
  "license_selection_required_at_submission": true,
  "dependencies": [
    "P01",
    "P03",
    "P05"
  ],
  "claim_ids": [
    "P06-C1",
    "P06-C2",
    "P06-C3",
    "P06-C4"
  ]
}