{
  "id": "WM-CDP-MANIFEST-001",
  "specification_id": "WM-CDP-SPEC-0.0.1",
  "benchmark_family": "WM-CDP",
  "name": "WholeMind Clinical Decision Benchmark Data Artifact Manifest",
  "version": "0.0.1",
  "status": "in_development",
  "status_label": "In development",
  "plain_language_name": "Library catalog",
  "human_description": "This file is the table of contents for the benchmark library. It tells a researcher or developer which data files define the current version of the benchmark and where to find them.",
  "description": "Top-level manifest for the versioned data artifacts that define the WholeMind Clinical Decision Benchmark.",
  "artifacts": [
    {
      "id": "WM-ART-001",
      "object_id": "WM-CDP-MANIFEST-001",
      "name": "Manifest JSON",
      "plain_language_name": "Library catalog",
      "version": "0.0.1",
      "status": "in_development",
      "path": "manifest.json",
      "human_description": "A directory of the benchmark artifacts. Use this first when building tools that need to know which files belong to the current specification.",
      "example_use": "A website or scoring app can load this file to discover the current rubric, failure modes, domains, scenario types, and scoring data structure."
    },
    {
      "id": "WM-ART-002",
      "object_id": "WM-CDP-RUBRIC-001",
      "name": "Rubric JSON",
      "plain_language_name": "Clinician scoring dimensions",
      "version": "0.0.1",
      "status": "in_development",
      "path": "rubric.json",
      "human_description": "The list of qualities clinicians score when judging whether a model response was a good next clinical move.",
      "example_use": "Display the eight scoring rows in a clinician rating form."
    },
    {
      "id": "WM-ART-003",
      "object_id": "WM-CDP-FAILURE-MODES-001",
      "name": "Failure Modes JSON",
      "plain_language_name": "Clinical error tags",
      "version": "0.0.1",
      "status": "in_development",
      "path": "failure-modes.json",
      "human_description": "The list of clinically meaningful ways a response can fail, even if it sounds warm or fluent.",
      "example_use": "Let clinicians tag whether a model reinforced reassurance seeking, missed an OCD cue, overreached diagnostically, or made another clinically important error."
    },
    {
      "id": "WM-ART-004",
      "object_id": "WM-CDP-CLINICAL-DOMAINS-001",
      "name": "Clinical Domains JSON",
      "plain_language_name": "Clinical topic areas",
      "version": "0.0.1",
      "status": "in_development",
      "path": "clinical-domains.json",
      "human_description": "The topic areas covered by the benchmark, such as school refusal, separation anxiety, OCD/reassurance seeking, and parent accommodation.",
      "example_use": "Group benchmark items and model scores by clinical area."
    },
    {
      "id": "WM-ART-005",
      "object_id": "WM-CDP-SCENARIO-TYPES-001",
      "name": "Scenario Types JSON",
      "plain_language_name": "Kinds of clinical decision points",
      "version": "0.0.1",
      "status": "in_development",
      "path": "scenario-types.json",
      "human_description": "The recurring kinds of situations the model is tested on, such as parent guidance, reassurance seeking, avoidance cycles, safety assessment, or repair after a poor response.",
      "example_use": "Filter benchmark items by what the model had to do at the decision point."
    },
    {
      "id": "WM-ART-006",
      "object_id": "WM-CDP-ANNOTATION-SCHEMA-001",
      "name": "Scoring Data Structure JSON",
      "plain_language_name": "Scoring data structure",
      "version": "0.0.1",
      "status": "in_development",
      "path": "annotation-schema.json",
      "human_description": "The data structure for one completed clinician score: which item was scored, which response was scored, who scored it, what ratings were given, which failure modes were tagged, and what notes were entered.",
      "example_use": "Define the output payload saved by a clinician scoring app."
    }
  ]
}