get_breakdown

Inputs

Types and defaults come from source annotations. Constraints described below still apply; the schema is not runtime SDK output.

slugRequired

text

detailOptional

text Default: "full"

{
  "type": "object",
  "properties": {
    "slug": {
      "type": "string"
    },
    "detail": {
      "type": "string",
      "default": "full"
    }
  },
  "required": [
    "slug"
  ]
}

Outputs and behavior

Fetch a studied video's Breakdown JSON by slug. Every response carries a state field: running | done | failed | absent. The job status is checked FIRST: while a (re-)study runs you get ready=false with the current phase; never superseded measurements (stale_breakdown_exists flags that an older snapshot is on disk). If the latest re-study failed but an earlier study succeeded, the last successful breakdown is served with state='failed' plus restudy_error, so you decide whether to judge it. A running job whose runner process died is reported as failed with recovery instructions. detail='summary' returns meta + pacing + counts without the per-word/per-caption arrays (use it for long videos).

The tool description documents its results and behavior. An exact output schema is not supplied; fields not described here are unknown.

What touches the network

This tool's description does not specify its network behavior. Tool-specific details are unknown.

Zing's network policy

Studying, rendering and memes run locally by default. Reference URL downloads require explicit consent. Requested model and font downloads are pinned by URL and SHA-256, cached locally and reported. Uoink is reached only through a peer you configure on your machine. Zing has no telemetry or analytics.

Source

Generated from the site's synced MCP inventory on every build. Descriptions and input schemas refer to this pinned source snapshot.

Source commit: d308dc3f9c2b42e305afd4d7d61e4bfa2f276bc4