Result objects

Understand identified products, candidate replacements, and batch results.

For a single product, Research Agent returns one structured result at the top level. A batch returns the same result objects inside researchResults.

Identified product

productBeingReplaced is Research Agent’s interpretation of the query.

Field	Type	Description
`manufacturer`	`string`	Product manufacturer.
`model`	`string`	Manufacturer model number or SKU.
`description`	`string`	Concise product description.
`specifications`	`string`	Technical details such as capacity, dimensions, finish, electrical requirements, and features.
`imageUrl`	`string`	Direct URL to a product image.

Candidate replacements

Each item in candidateReplacements is a potential substitute.

Field	Type	Description
`manufacturer`	`string`	Candidate manufacturer.
`model`	`string`	Manufacturer model number or SKU.
`description`	`string`	Concise candidate description.
`specifications`	`string`	Technical details used to evaluate the candidate.
`msrp`	`number \| null`	Estimated manufacturer suggested retail price when available.
`matchReasoning`	`string`	Plain-language explanation of the match and relevant differences.
`availableOnSibi`	`boolean`	Whether the candidate was confirmed as orderable on SIBI.
`imageUrl`	`string`	Direct URL to a product image.
`sibiProductId`	`string \| null`	SIBI product ID when the candidate was found on SIBI.
`sibiManufacturer`	`string \| null`	Manufacturer value returned by the SIBI catalog.
`sibiPrice`	`number \| null`	Current price returned by SIBI.
`sibiCheckoutLink`	`string \| null`	SIBI checkout URL for the candidate.

Batch results

When one task researches multiple products, the response contains a researchResults array:

{
  "status": "complete",
  "taskId": "<TASK_ID>",
  "taskUrl": "https://agents.sibi.ai/agents/<AGENT_ID>/tasks/<TASK_ID>",
  "researchResults": [
    {
      "label": "Kitchen refrigerator",
      "index": 1,
      "productBeingReplaced": {
        "manufacturer": "GE",
        "model": "GTS18GTHWW",
        "description": "Top-freezer refrigerator",
        "specifications": "...",
        "imageUrl": "https://example.com/source.jpg"
      },
      "candidateReplacements": []
    },
    {
      "label": "Laundry appliance",
      "index": 2,
      "error": "Could not identify the supplied model."
    }
  ]
}

index is the one-based position of the item in the original request. label identifies the item Research Agent processed. Results are ordered by index when an index is available.

A mixed batch can contain successful results and item-level errors while the task itself is complete. Check each entry for error before reading its product fields.

Failed tasks

A task-level failure returns status: "failed" and a top-level error:

{
  "status": "failed",
  "taskId": "<TASK_ID>",
  "taskUrl": "https://agents.sibi.ai/agents/<AGENT_ID>/tasks/<TASK_ID>",
  "error": "Task failed: <ERROR_MESSAGE>"
}