PlaygroundProxyRequest
JSON:API request envelope for the Playground proxy. The server forwards the operation described in `data.attributes` to the target `/v1` endpoint using the caller's own API key (decrypted server-side), so the key never reaches the browser.
Properties
| Field | Type | Description |
|---|---|---|
data* | object | Response body. Everything hangs off here, per the JSON:API envelope. |
type* | string | JSON:API resource type. Always `playground_request`. |
attributes* | object | The resource's own data, kept apart from its identity. |
mode* | string | `live` forwards the real request to the target endpoint and returns its response verbatim (passthrough). `playground` runs the scenario simulation engine and returns the same response the real endpoint would for the chosen scenario, with no side effects (no quota, webhooks, or notifications). |
method* | string | HTTP method of the target operation. One of `GET`, `POST`, `PUT`, `PATCH`, or `DELETE`. |
path* | string | Absolute target endpoint path, including the `/v1` prefix (for example `/v1/beneficiaries`). Must resolve to a documented operation; unlisted paths return `403 path_not_allowed`. |
query | object | Query parameters to append to the target URL. Optional. |
headers | object | Client headers to forward. Only a small allowlist is propagated (e.g. `Idempotency-Key`, `Accept-Language`); the rest are dropped. The target credential is set by the server — a client-supplied `Authorization` or `Cookie` is never forwarded. |
body | object | null | Request body for the target operation on non-`GET` methods. Optional / `null` when not applicable. |
scenario | object | Simulation scenario for `mode: playground` (ignored in `mode: live`). `id` picks the scenario (e.g. `valid`, `not_found`) from those listed by `GET /v1/playground/templates`. |
id | string | Identifier of the pre-loaded scenario. |
inputs | object | Inputs that parameterize the scenario. |
Used in operations
POST /v1/playground/proxy