Trigger prompt

Trigger scene query with a custom prompt

POST /api/scenequery/triggerPrompt

Trigger prompt

cURL

curl --request POST \
  --url https://api2.rhombussystems.com/api/scenequery/triggerPrompt \
  --header 'Content-Type: application/json' \
  --header 'x-auth-apikey: <api-key>' \
  --header 'x-auth-scheme: <x-auth-scheme>' \
  --data '
{
  "deviceFacetUuid": "AAAAAAAAAAAAAAAAAAAAAA.v0",
  "multiImageTimeDeltasSeconds": [
    -10,
    -5,
    5
  ],
  "prompt": "Count people in the scene",
  "region": {
    "coordinates": [
      {
        "x": 123,
        "y": 123
      }
    ]
  },
  "timestampMs": 1640995200000
}
'

200

{
  "error": true,
  "errorMsg": "<string>",
  "event": {
    "checkCondition": true,
    "image": "<string>",
    "imageOffsets": [
      123
    ],
    "prompt": "<string>",
    "timestampMs": 123,
    "value": "<string>"
  },
  "warningMsg": "<string>"
}

Authorizations

x-auth-apikey · string · header · required

Your Rhombus API key. Must be accompanied by the x-auth-scheme header set to api-token (or partner-api-token for partner endpoints).

Headers

x-auth-scheme · enum<string> · default: api-token · required

Authentication scheme identifier. Use api-token for standard API key authentication, partner-api-token for partner API key authentication. Must be paired with the x-auth-apikey header containing your API key.

Available options: api-token, api, partner-api-token, partner-api

Body

application/json

Request object for triggering a prompt.

deviceFacetUuid · string<DeviceFacetUuid> | null

RUUID with optional appended facet information

Example:

"AAAAAAAAAAAAAAAAAAAAAA.v0"

multiImageTimeDeltasSeconds · (integer<int32> | null)[] | null

Optional list of time deltas in seconds (relative to timestampMs) for multi-image mode. Max 5 entries, each in range [-30, 3600]. When non-empty, the prompt runs against a set of frames: the base image at timestampMs (always included) plus one frame per distinct non-zero delta. Duplicates and 0 are deduped. Omit or send null/empty for single-image mode.

Optional list of time deltas in seconds (relative to timestampMs) for multi-image mode. Max 5 entries, each in range [-30, 3600]. When non-empty, the prompt runs against a set of frames: the base image at timestampMs (always included) plus one frame per distinct non-zero delta. Duplicates and 0 are deduped. Omit or send null/empty for single-image mode.

Example:

[-10, -5, 5]

prompt · string | null

Prompt text to trigger

Example:

"Count people in the scene"

promptType · enum<string>

Available options: COUNT, PERCENT, BOOLEAN, TEXT

region · object

Show child attributes

timestampMs · integer<int64> | null

Timestamp in milliseconds for the trigger

Example:

1640995200000

Response

200 - application/json

OK

Response object for triggering a prompt.

error · boolean | null

errorMsg · string | null

event · object

Scene query report event from the triggered prompt

Show child attributes

warningMsg · string | null