POST
/
v1
/
teams
/
{teamId}
/
ai-media
/
inspections
TypeScript client
import { OctosparkClient } from '@octospark/sdk'

const client = new OctosparkClient({
  token: process.env.OCTOSPARK_TOKEN
})

const response = await client.aiMedia.inspectAsset({
  teamId: process.env.OCTOSPARK_TEAM_ID ?? "teamId",
  body: {},
})
{
  "assetId": "<string>",
  "toolCallId": "<string>",
  "model": "<string>",
  "transcript": {
    "lines": [
      "<string>"
    ],
    "fullText": "<string>"
  },
  "description": "<string>",
  "check": {
    "allMatched": true,
    "lines": [
      {
        "expected": "<string>",
        "closest": "<string>"
      }
    ]
  },
  "creditsUsedDecimillicents": 123
}

Authorizations

Authorization
string
header
required

Bearer authentication header of the form Bearer <token>, where <token> is your auth token.

Path Parameters

teamId
string
required

Team (workspace) id (UUID). Discover ids via the list teams endpoint or the agent context.

Body

application/json
assetId
string<uuid>
required

The team image asset to read back, e.g. a generated pin from getGeneration outputAssetIds.

Pattern: ^[0-9a-fA-F]{8}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{4}-[0-9a-fA-F]{12}$
idempotencyKey
string
required

Required. Reuse the same key with the same request after a lost response to get the original read-back instead of paying twice.

Required string length: 1 - 200
expectedText
string[]

Optional lines of text that should appear in the image. Each line is matched case- and punctuation-insensitively and reported as matched, mismatched (with the closest line actually read) or missing. The expected text is never shown to the vision model.

Maximum array length: 50

a string at most 500 character(s) long

Maximum string length: 500

Response

Success

assetId
string
required
toolCallId
string
required
model
string
required
transcript
object
required
description
string
required
check
object
required

Null unless expectedText was sent.

creditsUsedDecimillicents
number
required