Skip to main content
The analysis block carries derived data for retrieval and AI workflows. It is optional and discardable: a document without analysis is still canonical.

​
Shape

{
  "analysis": {
    "embeddings": { "<uuid>": [ /* embeddings */ ] },
    "passages":   [ /* passages */ ],
    "summaries":  [ /* summaries */ ],
    "settings":   { /* how this was produced */ }
  }
}

​
analysis.embeddings

A map keyed by the UUID of the target (scene, element, or character), whose value is an array of one or more embedding records: 
{
  "id":         "<embedding-uuid>",
  "model":      "text-embedding-3-large",
  "dimensions": 1536,
  "values":     [0.023, -0.041, ...],
  "source":     "text",
  "lang":       "en",
  "tokens":     420,
  "created":    "2026-01-14T10:30:00Z"
}
source enum: name · text · desc · heading · composite.

​
analysis.passages

Retrieval-sized text chunks keyed back to their source: 
{
  "id":       "<uuid>",
  "scene":    "<scene-uuid>",
  "elements": ["<uuid>", "<uuid>"],
  "text":     { "en": "..." },
  "tokens":   256,
  "overlap":  32
}

​
analysis.summaries

Document- or scene-scoped summaries: 
{
  "id":        "<uuid>",
  "scope":     "scene",
  "target":    "<scene-uuid>",
  "generated": true,
  "model":     "claude-opus-4-7",
  "text":      { "en": "..." },
  "created":   "2026-01-14T10:45:00Z"
}
scope enum: document · scene.

​
analysis.settings

Records how the analysis was produced, for reproducibility:
FieldTypeDescription
modelstring (≤100)Embedding or summary model identifier.
sizeinteger ≥1Chunk size. Default 512.
overlapinteger ≥0Chunk overlap in tokens. Default 64.
tokeniserstring (≤50)Tokeniser identifier (e.g. cl100k).
metameta mapFree-form string key/value extension.