Ritten OAUTH API

openapi: 3.2.0
info:
  title: External OAUTH API
  x-logo:
    url: https://storage.googleapis.com/ritten-ops-public-logos/rittenBanner
    backgroundColor: '#FFFFFF'
    altText: Ritten Logo
  description: "For Ritten Integrating Partners\n\n## Authentication\n\n- Request an access token with your provided integration credentials (`client_id` and `client_secret`) by calling our token endpoint:\n```bash\ncurl https://api.ritten.io/v1/oauth/token \\\n  -X POST \\\n  -H 'content-type: application/json' \\\n  -d '{\"client_id\":\"${client_id}\",\"client_secret\":\"${client_secret}\",\"audience\":\"https://external-api.ritten.io\",\"grant_type\":\"client_credentials\"}'\n```\n- Take the `access_token` from the response and use that as the `Bearer` token in your requests to our API.\n- Tokens are long-lived (24 hours / `expires_in: 86400`). The token endpoint also caches server-side, so rapid repeat calls won't hit Auth0 — but feel free to cache the access_token locally if you prefer.\n- The token endpoint itself does not require a Bearer token; the `client_secret` in the body is the authentication.\n\n> **Note:** When working in non-production environments, the API endpoints (and `audience` value) will be different.\n> For example, in the `beta` environment, the token endpoint is `https://api.beta.ritten.io/v1/oauth/token`\n> and the audience is `https://external-api.beta.ritten.io`.\n\n## Tenant Header\n\n- Make sure to add the tenant ID to the header of every request. This is the Ritten Clinic instance the request will target. Example:\n```\nX-Ritten-Tenant: ritclinic\n```\n\n## Rate Limiting\n\nTwo layers of rate limiting apply: per-request limits on API calls, and per-app limits on token minting.\n\n### API request rate limit\n\nApplied to authenticated API calls (everything except `/v1/oauth/token`):\n\n- 50 requests per second sustained rate\n- 100 requests burst allowance\n\nYou can make up to 100 requests in a short burst, but over time your average must stay at or below 50 requests per second. Think of it as a bucket that holds 100 tokens and refills at 50 tokens per second. Each request consumes one token. You'll receive a `429 Too Many Requests` response when this is triggered.\n\n### Token mint quota (Auth0)\n\nA separate per-application limit on how often you can mint new access tokens:\n\n- 2 mints per hour\n- 3 mints per day\n\nThese limits are applied at the Auth0 layer and count mints across both the legacy direct path and the cached `/v1/oauth/token` endpoint combined. **The cached endpoint is designed so that one mint per day is sufficient for any traffic volume** — the proxy serves all subsequent requests from the cached token. If you migrate to the cached endpoint, you will not notice these limits.\n\nToken mint quotas currently apply to all newly-provisioned integrator clients. They will be rolled out to existing clients on a separate schedule, and you will be contacted before that change applies to you.\n"
  version: 1.0.0
servers:
- url: https://api.ritten.io/v1
tags:
- name: oauth
  description: 'OAuth 2.0 token endpoint for obtaining access tokens. This is the

    recommended way to authenticate with the Ritten External API.

    '
paths:
  /oauth/token:
    post:
      tags:
      - oauth
      summary: Obtain an access token (OAuth 2.0 client_credentials)
      description: 'Exchanges integrator credentials for a 24-hour access token to use as a `Bearer`

        token on subsequent API calls. This endpoint is **unauthenticated** at the gateway

        layer — your `client_secret` in the request body is the authentication. Ritten

        forwards the credentials to Auth0, validates the response, and caches the resulting

        token server-side so repeated calls do not consume your Auth0 mint quota.


        The response shape mirrors the OAuth 2.0 / Auth0 `/oauth/token` response so existing

        OAuth2 client libraries work without modification.


        **Mint quota interaction:** the per-app token mint quota (2/hour, 3/day) counts

        mints actually performed against Auth0. Because this endpoint caches server-side,

        repeated calls within a 24-hour window typically result in zero additional Auth0

        mints — so calling here is far cheaper against your quota than calling Auth0

        directly. See the Authentication and Rate Limiting sections above.

        '
      operationId: postOAuthToken
      security: []
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/OAuthTokenRequest'
          application/x-www-form-urlencoded:
            schema:
              $ref: '#/components/schemas/OAuthTokenRequest'
      responses:
        200:
          description: Access token issued.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/OAuthTokenResponse'
        400:
          description: 'Malformed request. The `error` field will be one of `invalid_request`,

            `unsupported_grant_type`, or `invalid_audience`.

            '
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/OAuthErrorResponse'
        401:
          description: 'Auth0 rejected the supplied credentials (`error: invalid_client` or similar).

            '
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/OAuthErrorResponse'
        429:
          description: 'Rate limit exceeded. Either the per-IP request rate limit on this endpoint,

            or the per-app Auth0 mint quota (2/hour, 3/day) has been reached. Retry

            after a short delay. If 429s persist, inspect whether you are hitting

            per-IP request limits vs. triggering fresh Auth0 mints, then reach out

            to Ritten for support.

            '
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/OAuthErrorResponse'
        502:
          description: 'Auth0 was unreachable or returned a 5xx error. Retry — Ritten does not

            cache failed responses.

            '
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/OAuthErrorResponse'
components:
  schemas:
    OAuthTokenRequest:
      type: object
      required:
      - grant_type
      - client_id
      - client_secret
      - audience
      properties:
        grant_type:
          type: string
          enum:
          - client_credentials
          description: OAuth 2.0 grant type. Must be `client_credentials`.
          example: client_credentials
        client_id:
          type: string
          description: Your Auth0 M2M client ID.
        client_secret:
          type: string
          format: password
          writeOnly: true
          description: Your Auth0 M2M client secret.
        audience:
          type: string
          description: 'The audience for the requested token. Must equal the env-specific

            external-api audience (e.g. `https://external-api.ritten.io` in

            production, `https://external-api.beta.ritten.io` in beta).

            '
          example: https://external-api.ritten.io
    OAuthErrorResponse:
      type: object
      required:
      - error
      properties:
        error:
          type: string
          description: 'OAuth 2.0 error code. Common values: `invalid_request`,

            `unsupported_grant_type`, `invalid_audience`, `invalid_client`,

            `bad_gateway`, `rate_limit_exceeded`.

            '
        error_description:
          type: string
          description: Human-readable explanation of the error.
    OAuthTokenResponse:
      type: object
      required:
      - access_token
      - token_type
      - expires_in
      properties:
        access_token:
          type: string
          description: The access token to use as a `Bearer` token on subsequent API calls.
        token_type:
          type: string
          example: Bearer
        expires_in:
          type: integer
          format: int64
          description: Token lifetime in seconds (currently 86400 / 24h).
          example: 86400
        scope:
          type: string
          description: Space-separated list of granted scopes (may be empty).