openapi: 3.0.3
info:
  title: Runloop Account Scenario API
  description: Representative OpenAPI description of the Runloop API - AI-native cloud development environments (devboxes) and an agent benchmarking platform. Endpoints cover devbox lifecycle and command execution, blueprints, disk snapshots, code mounts / repositories (via CodeMount configuration), scenarios, benchmarks and objects. All requests are authenticated with a Bearer API key. This is a faithful representative subset of the full Stainless-generated Runloop specification.
  termsOfService: https://www.runloop.ai/
  contact:
    name: Runloop Support
    url: https://docs.runloop.ai/
  version: '1.0'
servers:
- url: https://api.runloop.ai
  description: Runloop production API
security:
- bearerAuth: []
tags:
- name: Scenario
  description: Define and run coding-agent evaluation scenarios.
paths:
  /v1/scenarios:
    post:
      operationId: createScenario
      tags:
      - Scenario
      summary: Create a Scenario
      requestBody:
        required: true
        content:
          application/json:
            schema:
              $ref: '#/components/schemas/ScenarioCreateParameters'
      responses:
        '200':
          description: The created Scenario.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ScenarioView'
    get:
      operationId: listScenarios
      tags:
      - Scenario
      summary: List Scenarios
      parameters:
      - $ref: '#/components/parameters/limit'
      - $ref: '#/components/parameters/startingAfter'
      responses:
        '200':
          description: A page of Scenarios.
          content:
            application/json:
              schema:
                type: object
                properties:
                  scenarios:
                    type: array
                    items:
                      $ref: '#/components/schemas/ScenarioView'
                  has_more:
                    type: boolean
  /v1/scenarios/{id}:
    get:
      operationId: getScenario
      tags:
      - Scenario
      summary: Get a Scenario
      parameters:
      - $ref: '#/components/parameters/idPath'
      responses:
        '200':
          description: The requested Scenario.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ScenarioView'
  /v1/scenarios/start_run:
    post:
      operationId: startScenarioRun
      tags:
      - Scenario
      summary: Start a new ScenarioRun
      requestBody:
        required: true
        content:
          application/json:
            schema:
              type: object
              required:
              - scenario_id
              properties:
                scenario_id:
                  type: string
                run_name:
                  type: string
                metadata:
                  type: object
                  additionalProperties:
                    type: string
      responses:
        '200':
          description: The started ScenarioRun.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ScenarioRunView'
  /v1/scenarios/runs/{id}:
    get:
      operationId: getScenarioRun
      tags:
      - Scenario
      summary: Get a ScenarioRun
      parameters:
      - $ref: '#/components/parameters/idPath'
      responses:
        '200':
          description: The requested ScenarioRun.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ScenarioRunView'
  /v1/scenarios/runs/{id}/score:
    post:
      operationId: scoreScenarioRun
      tags:
      - Scenario
      summary: Score a ScenarioRun
      parameters:
      - $ref: '#/components/parameters/idPath'
      responses:
        '200':
          description: The scored ScenarioRun.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ScenarioRunView'
  /v1/scenarios/runs/{id}/complete:
    post:
      operationId: completeScenarioRun
      tags:
      - Scenario
      summary: Complete a ScenarioRun
      parameters:
      - $ref: '#/components/parameters/idPath'
      responses:
        '200':
          description: The completed ScenarioRun.
          content:
            application/json:
              schema:
                $ref: '#/components/schemas/ScenarioRunView'
components:
  parameters:
    limit:
      name: limit
      in: query
      required: false
      schema:
        type: integer
      description: Number of items to return.
    startingAfter:
      name: starting_after
      in: query
      required: false
      schema:
        type: string
      description: Cursor for pagination; return items after this id.
    idPath:
      name: id
      in: path
      required: true
      schema:
        type: string
  schemas:
    ScenarioView:
      type: object
      properties:
        id:
          type: string
        name:
          type: string
        input_context:
          type: object
        scoring_contract:
          type: object
        reference_output:
          type: string
        is_public:
          type: boolean
        metadata:
          type: object
          additionalProperties:
            type: string
    ScenarioRunView:
      type: object
      properties:
        id:
          type: string
        name:
          type: string
        scenario_id:
          type: string
        devbox_id:
          type: string
        benchmark_run_id:
          type: string
        state:
          type: string
          enum:
          - running
          - scoring
          - scored
          - completed
          - canceled
        scoring_contract_result:
          type: object
        start_time_ms:
          type: integer
          format: int64
        duration_ms:
          type: integer
          format: int64
        metadata:
          type: object
          additionalProperties:
            type: string
    ScenarioCreateParameters:
      type: object
      required:
      - name
      properties:
        name:
          type: string
        input_context:
          type: object
        environment_parameters:
          type: object
        scoring_contract:
          type: object
          description: Definition of the scorers used to grade a run of this scenario.
        reference_output:
          type: string
        metadata:
          type: object
          additionalProperties:
            type: string
  securitySchemes:
    bearerAuth:
      type: http
      scheme: bearer
      description: 'Runloop API key sent as: Authorization Bearer <API_KEY>.'