Advanced Micro Devices Agentic Access

x-agentic-access generated

Advanced Micro Devices exposes 13 API operations that an AI agent could call, of which 3 are state-changing ‘acting’ operations. This is a recommended x-agentic-access execution contract — the scope, audience, consequence tier, short-lived token constraints, and escalation each action should carry before it is handed to an autonomous agent.

By consequence: 10 read, 1 write, 1 physical, and 1 safety-critical.

1 operation are classed safety-critical and should require human-in-the-loop approval at runtime.

Contracts are classified heuristically from the provider’s OpenAPI and refresh on every APIs.io network build; audience is bound per deployment. The model follows Curity’s Access Intelligence (apidays Munich 2026). Browse every provider’s agent contracts at agentic-access.apis.io.

AICloud ComputingGPUHPCMachine LearningSemiconductorFortune 500

Operations: 13 Acting: 3 Human-in-the-loop: 1 Method: generated

By consequence

read 10 write 1 physical 1 safety-critical 1

Highest-consequence actions

The physical and safety-critical operations an agent could invoke — the ones that most warrant scoped tokens, tight TTLs, and escalation. Full per-operation contracts are in the source below.

Method	Path	Consequence	Human-in-loop
DELETE	/instances/{instanceId}	safety-critical	required
POST	/models	physical	conditional

Source

Agentic Access

generated: '2026-07-15'
method: generated
source: openapi/amd-developer-cloud-api-openapi.yml, openapi/amd-rocm-management-api-openapi.yml
description: Recommended x-agentic-access execution contracts, classified heuristically from
  the OpenAPI. A governance starting point for exposing this API to AI agents — review and bind
  audience per deployment. See research/curity/agentic-governance/.
summary:
  operations: 13
  by_action_class:
    connected: 10
    acting: 3
  by_consequence:
    read: 10
    write: 1
    safety-critical: 1
    physical: 1
  human_in_the_loop_required: 1
operations:
- path: /instances
  method: get
  operationId: listInstances
  x-agentic-access:
    action-class: connected
    consequence: read
    subject: optional
    token:
      max-ttl: 3600
    audit: none
- path: /instances
  method: post
  operationId: createInstance
  x-agentic-access:
    action-class: acting
    consequence: write
    subject: required
    audience: null
    token:
      max-ttl: 900
    escalation:
      human-in-the-loop: conditional
      triggers:
      - abnormal
      - high-value
    audit: required
- path: /instances/{instanceId}
  method: get
  operationId: getInstance
  x-agentic-access:
    action-class: connected
    consequence: read
    subject: optional
    token:
      max-ttl: 3600
    audit: none
- path: /instances/{instanceId}
  method: delete
  operationId: terminateInstance
  x-agentic-access:
    action-class: acting
    consequence: safety-critical
    subject: required
    audience: null
    token:
      max-ttl: 120
      exchange: true
      purpose-required: true
      proof-of-possession: true
    escalation:
      human-in-the-loop: required
    audit: required
- path: /instances/{instanceId}/metrics
  method: get
  operationId: getInstanceMetrics
  x-agentic-access:
    action-class: connected
    consequence: read
    subject: optional
    token:
      max-ttl: 3600
    audit: none
- path: /models
  method: get
  operationId: listModels
  x-agentic-access:
    action-class: connected
    consequence: read
    subject: optional
    token:
      max-ttl: 3600
    audit: none
- path: /models
  method: post
  operationId: deployModel
  x-agentic-access:
    action-class: acting
    consequence: physical
    subject: required
    audience: null
    token:
      max-ttl: 300
      exchange: true
      purpose-required: true
    escalation:
      human-in-the-loop: conditional
      triggers:
      - abnormal
      - high-value
    audit: required
- path: /credits
  method: get
  operationId: getCredits
  x-agentic-access:
    action-class: connected
    consequence: read
    subject: optional
    token:
      max-ttl: 3600
    audit: none
- path: /devices
  method: get
  operationId: listDevices
  x-agentic-access:
    action-class: connected
    consequence: read
    subject: optional
    token:
      max-ttl: 3600
    audit: none
- path: /devices/{deviceId}
  method: get
  operationId: getDevice
  x-agentic-access:
    action-class: connected
    consequence: read
    subject: optional
    token:
      max-ttl: 3600
    audit: none
- path: /devices/{deviceId}/health
  method: get
  operationId: getDeviceHealth
  x-agentic-access:
    action-class: connected
    consequence: read
    subject: optional
    token:
      max-ttl: 3600
    audit: none
- path: /devices/{deviceId}/performance
  method: get
  operationId: getDevicePerformance
  x-agentic-access:
    action-class: connected
    consequence: read
    subject: optional
    token:
      max-ttl: 3600
    audit: none
- path: /system/rocm-version
  method: get
  operationId: getRocmVersion
  x-agentic-access:
    action-class: connected
    consequence: read
    subject: optional
    token:
      max-ttl: 3600
    audit: none