schema

$limai schema <subcommand>

Inspect and update the extraction schema of a model, and read its extracted data.

Options

--jsonOutput raw JSON

--verboseEnable debug logging

Every subcommand takes one ID, which may be a model ID, the ID of a deployment that points at a model, or a model's extraction schema ID.

Subcommands

`limai schema get <deploymentId>`

Get the details of a model's extraction schema, including its tables and columns.

limai schema get <deploymentId>

`limai schema update <deploymentId>`

Update extraction schema settings. Pass at least one of the flags below, or the command is rejected.

$limai schema update <deploymentId>

Update extraction schema settings.

Options

--body <file>JSON file with schema settings to update

--dynamic-mapping <on|off>Enable or disable dynamic mapping (turns big-table extraction off when enabled)

--bbox-tier <tier>Bounding box tier: OFF, TEXT, FAST, BALANCED or PRO

--reasoning-level <level>Reasoning depth: LOW, MEDIUM or HIGH

Configuration is edited on the model. Addressing this command with a deployment's ID is rejected with 409 POINTER_NOT_WRITABLE.

`limai schema data <deploymentId>`

Get all extracted data for a model.

limai schema data <deploymentId>

`--include metadata`

--include takes a comma-separated list of enrichments; metadata is the only accepted value, and any other value is rejected before the request leaves the CLI.

limai schema data <deploymentId> --include metadata

With metadata, every top-level row — the object for a single-row file, each array element for a multi-row file — carries a _meta block describing the document it came from:

"_meta": {
  "documentId": "cmm3k7r2v0007",
  "fileName": "broker-quote-2026-09.pdf",
  "submittedAt": "2026-09-01T10:12:33.000Z",
  "submittedBy": { "id": "usr_8a2f", "name": "Jane Doe", "email": "jane@acme.com" },
  "extractedAt": "2026-09-01T10:14:02.000Z"
}

Field	Description
`documentId`	The file's ID.
`fileName`	The uploaded file name.
`submittedAt`	When the file was uploaded.
`submittedBy`	The user who uploaded the document. `null` when the uploader is unknown.
`extractedAt`	When the latest extraction run completed. `null` when no run has completed.

--json prints the raw payload with _meta included. Plain table mode prints a summary table — Filename, Submitted by, Submitted at, Extracted at, Document ID, one row per file — ahead of the usual key/value dump.

Examples

View schema structure

limai schema get dep_abc123

Export all extracted data as JSON

limai schema data dep_abc123 --json

Export extracted data with document metadata

limai schema data dep_abc123 --include metadata --json

schema

On this page