Triton Inference Server NVIDIA Triton Inference Server HTTP/REST API — Model Repository

naftiko: 1.0.0-alpha2
info:
  label: Triton Inference Server NVIDIA Triton Inference Server HTTP/REST API — Model Repository
  description: 'Triton Inference Server NVIDIA Triton Inference Server HTTP/REST API — Model Repository. 3 operations. Lead
    operation: Triton Inference Server List Models in the Repository. Self-contained Naftiko capability covering one Triton
    business surface.'
  tags:
  - Triton
  - Model Repository
  created: '2026-05-19'
  modified: '2026-05-19'
binds:
- namespace: env
  keys:
    TRITON_API_KEY: TRITON_API_KEY
capability:
  consumes:
  - type: http
    namespace: http-rest-model-repository
    baseUri: http://localhost:8000
    description: Triton Inference Server NVIDIA Triton Inference Server HTTP/REST API — Model Repository business capability.
      Self-contained, no shared references.
    resources:
    - name: v2-repository-index
      path: /v2/repository/index
      operations:
      - name: repositoryindex
        method: POST
        description: Triton Inference Server List Models in the Repository
        outputRawFormat: json
        outputParameters:
        - name: result
          type: object
          value: $.
        inputParameters:
        - name: body
          in: body
          type: object
          description: Request body (JSON).
          required: false
    - name: v2-repository-models-model_name-load
      path: /v2/repository/models/{model_name}/load
      operations:
      - name: modelload
        method: POST
        description: Triton Inference Server Load or Reload a Model
        outputRawFormat: json
        outputParameters:
        - name: result
          type: object
          value: $.
        inputParameters:
        - name: body
          in: body
          type: object
          description: Request body (JSON).
          required: false
    - name: v2-repository-models-model_name-unload
      path: /v2/repository/models/{model_name}/unload
      operations:
      - name: modelunload
        method: POST
        description: Triton Inference Server Unload a Model
        outputRawFormat: json
        outputParameters:
        - name: result
          type: object
          value: $.
        inputParameters:
        - name: body
          in: body
          type: object
          description: Request body (JSON).
          required: false
  exposes:
  - type: rest
    namespace: http-rest-model-repository-rest
    port: 8080
    description: REST adapter for Triton Inference Server NVIDIA Triton Inference Server HTTP/REST API — Model Repository.
      One Spectral-compliant resource per consumed operation, prefixed with /v1.
    resources:
    - path: /v1/v2/repository/index
      name: v2-repository-index
      description: REST surface for v2-repository-index.
      operations:
      - method: POST
        name: repositoryindex
        description: Triton Inference Server List Models in the Repository
        call: http-rest-model-repository.repositoryindex
        with:
          body: rest.body
        outputParameters:
        - type: object
          mapping: $.
    - path: /v1/v2/repository/models/{model-name}/load
      name: v2-repository-models-model-name-load
      description: REST surface for v2-repository-models-model_name-load.
      operations:
      - method: POST
        name: modelload
        description: Triton Inference Server Load or Reload a Model
        call: http-rest-model-repository.modelload
        with:
          body: rest.body
        outputParameters:
        - type: object
          mapping: $.
    - path: /v1/v2/repository/models/{model-name}/unload
      name: v2-repository-models-model-name-unload
      description: REST surface for v2-repository-models-model_name-unload.
      operations:
      - method: POST
        name: modelunload
        description: Triton Inference Server Unload a Model
        call: http-rest-model-repository.modelunload
        with:
          body: rest.body
        outputParameters:
        - type: object
          mapping: $.
  - type: mcp
    namespace: http-rest-model-repository-mcp
    port: 9090
    transport: http
    description: MCP adapter for Triton Inference Server NVIDIA Triton Inference Server HTTP/REST API — Model Repository.
      One tool per consumed operation, routed inline through this capability's consumes block.
    tools:
    - name: triton-inference-server-list-models
      description: Triton Inference Server List Models in the Repository
      hints:
        readOnly: true
        destructive: false
        idempotent: false
      call: http-rest-model-repository.repositoryindex
      with:
        body: tools.body
      outputParameters:
      - type: object
        mapping: $.
    - name: triton-inference-server-load-reload
      description: Triton Inference Server Load or Reload a Model
      hints:
        readOnly: false
        destructive: false
        idempotent: false
      call: http-rest-model-repository.modelload
      with:
        body: tools.body
      outputParameters:
      - type: object
        mapping: $.
    - name: triton-inference-server-unload-model
      description: Triton Inference Server Unload a Model
      hints:
        readOnly: false
        destructive: false
        idempotent: false
      call: http-rest-model-repository.modelunload
      with:
        body: tools.body
      outputParameters:
      - type: object
        mapping: $.