---
title: "List AI Gateway Auth Strategy Usage"
method: GET
path: "/v1/ai-gateways/{gatewayId}/auth-strategies/{authStrategyIdOrName}/usage"
tags: ["AI Gateway Auth Strategies"]
---

# List AI Gateway Auth Strategy Usage

`GET /v1/ai-gateways/{gatewayId}/auth-strategies/{authStrategyIdOrName}/usage`

## Request

- Base URL: `https://global.api.konghq.com`
- URL: `https://global.api.konghq.com/v1/ai-gateways/{gatewayId}/auth-strategies/{authStrategyIdOrName}/usage`
- Auth: one of:
  - HTTP bearer
  - HTTP bearer
  - HTTP bearer
  - HTTP bearer

Returns the entities that reference a specific AI Gateway auth strategy, so dependencies can be identified before the auth strategy is deleted.

Each entity list is limited to the first N records per entity type.

## Response `200`

A successful response listing usage of an Auth Strategy.

- object
  - `agents` AIGatewayAgent[], required
    - `display_name` string, required: The display name for this agent.
    - `name` string, required: Identifier for an AI Gateway entity. In some cases, this may be the entity name or ID.
    - `enabled` boolean: Whether the Agent is enabled.
    - `type` 'a2a' | 'http', required: The type of the agent.
    - `policies` string[]: List of policy references.
    - `access` AIGatewayAgentAccess: Access control configuration for an agent.
      - `acls` union: Access control rules. Configure exactly one of `allow`, `deny`, `allow_when`, or `deny_when`.
        - AIGatewayAllowACL
          - `allow` string[], required: List of Consumer Groups Names, or Authenticated Groups Names that are permitted access.
        - AIGatewayDenyACL
          - `deny` string[], required: List of Consumer Groups Names, or Authenticated Groups Names that are denied access.
        - AIGatewayAllowWhenACL
          - `allow_when` string[], required: Allow the request if it matches any of these CEL boolean expressions evaluated against the request context (consumer, principal, HTTP attributes, consumer groups, etc.). Exactly one of `allow`, `deny`, `allow_when`, or `deny_when` must be specified. **Requires a minimum runtime version of `2.2`**.
        - AIGatewayDenyWhenACL
          - `deny_when` string[], required: Deny the request if it matches any of these CEL boolean expressions evaluated against the request context (consumer, principal, HTTP attributes, consumer groups, etc.). Exactly one of `allow`, `deny`, `allow_when`, or `deny_when` must be specified. **Requires a minimum runtime version of `2.2`**.
      - `identity_providers` AIGatewayIdentityProviderReference[]: List of identity providers for granting access to the agent. At most 1 identity provider of each identity provider type can be referenced. Deprecated: use `auth_strategies` instead. The two are mutually exclusive.
      - `auth_strategies` AIGatewayAuthStrategyReference[]: List of auth strategies for granting access to the agent. At most 1 auth strategy of each auth strategy type can be referenced.
    - `config` object, required: Configuration for the agent. The structure varies depending on the agent type.
      - `url` string, uri, required: Helper field to set protocol, host, port and path of the upstream A2A Agent using a URL. This is the same as a Kong Gateway Service URL: ${scheme}://${host}:${port}/${path}
      - `upstream` AIGatewayUpstreamConfig: Configuration applied when proxying to the upstream service, including authentication.
        - `auth` AIGatewayUpstreamAuthAWS: AWS IAM (SigV4) authentication for the upstream service.
          - `type` 'aws', required
          - `access_key_id` string: The access key id for authenticating with static IAM User credentials. This field is [referenceable](https://developer.konghq.com/gateway/entities/vault/#how-do-i-reference-secrets-stored-in-a-vault).
          - `secret_access_key` string: The secret access key for authenticating with static IAM User credentials. This field is [referenceable](https://developer.konghq.com/gateway/entities/vault/#how-do-i-reference-secrets-stored-in-a-vault).
          - `session_token` string: The session token for authenticating with temporary IAM credentials. This field is [referenceable](https://developer.konghq.com/gateway/entities/vault/#how-do-i-reference-secrets-stored-in-a-vault).
          - `region` string: The AWS region of the upstream service. Overrides the region inferred from the environment.
          - `assume_role_arn` string: The ARN of the IAM role to assume for generating authentication tokens. This field is [referenceable](https://developer.konghq.com/gateway/entities/vault/#how-do-i-reference-secrets-stored-in-a-vault).
          - `role_session_name` string: The session name for the temporary credentials when assuming the IAM role. This field is [referenceable](https://developer.konghq.com/gateway/entities/vault/#how-do-i-reference-secrets-stored-in-a-vault).
          - `sts_endpoint_url` string: The STS endpoint URL to use for generating authentication tokens. If not specified, the default AWS STS endpoint will be used.
      - `route` AIGatewayRouteConfig: Configuration for an AI Gateway route.
        - `headers` object: One or more lists of values indexed by header name that will cause this route to match if present in the request. The `Host` header cannot be used with this attribute: hosts should be specified using the `hosts` attribute. When `headers` contains only one value and that value starts with the special prefix `~*`, the value is interpreted as a regular expression.
        - `hosts` string[]: A list of domain names that match this route. Note that the hosts value is case sensitive.
        - `https_redirect_status_code` integer: The status code Kong responds with when all properties of a route match except the protocol i.e. if the protocol of the request is `HTTP` instead of `HTTPS`. `Location` header is injected by Kong if the field is set to 301, 302, 307 or 308. Note: This config applies only if the route is configured to only accept the `https` protocol.
        - `methods` string[]: A list of HTTP methods that match this route.
        - `paths` string[]: A list of paths that match this route.
        - `preserve_host` boolean: When matching a route via one of the `hosts` domain names, use the request `Host` header in the upstream request headers. If set to `false`, the upstream `Host` header will be that of the service's `host`.
        - `protocols` string[]: An array of the protocols this route should allow. See the [route Object](#route-object) section for a list of accepted protocols. When set to only `https`, HTTP requests are answered with an upgrade error. When set to only `http`, HTTPS requests are answered with an error.
        - `regex_priority` integer: A number used to choose which route resolves a given request when several routes match it using regexes simultaneously. When two routes match the path and have the same `regex_priority`, the older one (lowest `created_at`) is used. Note that the priority for non-regex routes is different (longer non-regex routes are matched before shorter ones).
        - `request_buffering` boolean: Whether to enable request body buffering or not. With HTTP 1.1, it may make sense to turn this off on services that receive data with chunked transfer encoding.
        - `response_buffering` boolean: Whether to enable response body buffering or not. With HTTP 1.1, it may make sense to turn this off on services that send data with chunked transfer encoding.
        - `strip_path` boolean: When matching a route via one of the `paths`, strip the matching prefix from the upstream request URL.
        - `tags` string[]: An optional set of strings associated with the route for grouping and filtering.
      - `proxy` AIGatewayProxyConfig: HTTP/HTTPS proxy configuration for outbound requests to the upstream AI provider.
        - `http_proxy` object: HTTP proxy server to route plaintext outbound requests through.
          - `host` string: A string representing a host name, such as example.com.
          - `port` integer: An integer representing a port number between 0 and 65535, inclusive.
        - `https_proxy` object: HTTPS proxy server to route TLS outbound requests through.
          - `host` string: A string representing a host name, such as example.com.
          - `port` integer: An integer representing a port number between 0 and 65535, inclusive.
        - `proxy_scheme` 'http': The proxy scheme to use when connecting to the proxy server.
        - `auth` object: Credentials used to authenticate to the proxy server.
          - `username` string: The username to use for proxy authentication. This field is [referenceable](https://developer.konghq.com/gateway/entities/vault/#how-do-i-reference-secrets-stored-in-a-vault).
          - `password` string: The password to use for proxy authentication. This field is [referenceable](https://developer.konghq.com/gateway/entities/vault/#how-do-i-reference-secrets-stored-in-a-vault).
        - `no_proxy` string: Comma-separated list of hosts that should not be proxied.
      - `max_request_body_size` integer: Maximum size of request body to parse. Set to 0 for unlimited.
      - `logging` object: Configuration for AI Gateway logging.
        - `payloads` boolean
        - `max_payload_size` integer: Maximum size in bytes for logged request/response payloads. Payloads exceeding this size will be truncated.
      - `certificates` AIGatewayCertificateReference[]: Client certificates used for mutual TLS (mTLS) authentication with the upstream agent endpoint, referenced by name. Each entry must be the `name` of an existing AI Certificate on the same AI Gateway. The certificates are applied to the Kong Gateway Service generated for this Agent, which proxies requests to `config.url`. The generated Service presents a single client certificate (`client_certificate`) during the TLS handshake with the upstream, so at most one certificate may be configured. If unset, no client certificate is presented to the upstream.
      - `ca_certificates` AIGatewayCACertificateReference[]: CA certificates used to verify the TLS certificate of the upstream agent endpoint, referenced by name. Each entry must be the `name` of an existing AI CA Certificate on the same AI Gateway. The CA certificates build the trust store (`ca_certificates`) of the Kong Gateway Service generated for this Agent, which proxies requests to `config.url`. If unset, the default trust store is used to verify the upstream's TLS certificate.
    - `labels` PublicLabels: Public labels store information about an entity that can be used for filtering a list of objects. Public labels are intended to store **PUBLIC** metadata. Keys must be of length 1-63 characters, and cannot start with "kong", "konnect", "mesh", "kic", or "_".
    - `managed_by` ManagedBy: Stores information about what manages this entity, such as the tool or system responsible for its lifecycle (for example, `terraform`). Keys must be 1–63 characters long, start and end with an alphanumeric character, and may contain dashes (`-`), underscores (`_`), and dots (`.`). Values must be 1–127 characters long and must not be empty. Keys and values must not contain spaces or control characters.
    - `id` string, uuid, required: Contains a unique identifier used for this resource.
    - `created_at` string, date-time, required: An ISO-8601 timestamp representation of entity creation date.
    - `updated_at` string, date-time, required: An ISO-8601 timestamp representation of entity update date.
  - `models` AIGatewayModel[], required
    - union: Configuration for an AI Gateway model.
      - object: Configuration for proxying asynchronous requests/responses to/from an AI Gateway model using the files, batches, and skills APIs.
        - `id` string, uuid, required: Contains a unique identifier used for this resource.
        - `created_at` string, date-time, required: An ISO-8601 timestamp representation of entity creation date.
        - `updated_at` string, date-time, required: An ISO-8601 timestamp representation of entity update date.
        - `display_name` string, required: The display name for this model instance.
        - `name` string, required: Identifier for an AI Gateway entity. In some cases, this may be the entity name or ID.
        - `enabled` boolean: Whether the model is enabled.
        - `access` AIGatewayModelAccess: Access control configuration for a model.
          - `acls` union: Access control rules. Configure exactly one of `allow`, `deny`, `allow_when`, or `deny_when`.
            - AIGatewayAllowACL
              - …
            - AIGatewayDenyACL
              - …
            - AIGatewayAllowWhenACL
              - …
            - AIGatewayDenyWhenACL
              - …
          - `identity_providers` AIGatewayIdentityProviderReference[]: List of identity providers for granting access to the model. At most 1 identity provider of each identity provider type can be referenced. Deprecated: use `auth_strategies` instead. The two are mutually exclusive.
          - `auth_strategies` AIGatewayAuthStrategyReference[]: List of auth strategies for granting access to the model. At most 1 auth strategy of each auth strategy type can be referenced.
        - `formats` AIGatewayModelFormat[], required: List of request/response formats supported by this model.
          - `type` '*' | 'anthropic' | 'bedrock' | 'cohere' | 'gemini' | 'huggingface' | 'openai' | 'passthrough' | 'typesafe': The request and response format this model supports. Set to a specific format to restrict the model to that format only, or to `*` to allow all formats that the targets support at the gateway's min_runtime_version. A format that needs a newer min_runtime_version is left out of `*`. The `*` format never includes passthrough. **`passthrough` requires a minimum runtime version of `2.2`**. **`typesafe` requires a minimum runtime version of `2.2`**.
        - `targets` AIGatewayTarget[], required: One or more backend models that this model entry routes to.
          - `name` string, required: The name of the model defined in the upstream provider that will be executed.
          - `weight` integer: The weight this target gets within the upstream load balancer
          - `semantic_description` string: The semantic description of the target, required if using semantic load balancing. Specially, setting this to 'CATCHALL' will indicate such target to be used when no other targets match the semantic threshold.
          - `allow_auth_override` boolean: When enabled, request-level auth parameters (such as API keys or bearer tokens) will override the static values defined for the provider.
          - `provider` string, required: Reference to a model provider instance by name.
          - `config` union, required: Configuration for a target model.
            - AIGatewayTargetAnthropicConfig: Anthropic-specific configuration for a model.
              - …
            - AIGatewayTargetAzureConfig: Azure-specific configuration for a model.
              - …
            - AIGatewayTargetBasetenConfig: Baseten-specific configuration for a model. Requests are sent to Baseten Model APIs at `https://inference.baseten.co/v1` by default, and the target `name` is the Baseten model slug (for example, `deepseek-ai/DeepSeek-V4-Pro-0813`). To route to a dedicated Baseten deployment, set `upstream_url` to the deployment's full OpenAI-compatible endpoint (for example, `https://model-{model_id}.api.baseten.co/environments/{environment}/sync/v1/chat/completions`) or use passthrough format. **Requires a minimum runtime version of `2.3`**.
              - …
            - AIGatewayTargetBedrockConfig: AWS Bedrock-specific configuration for a model.
              - …
            - AIGatewayTargetCerebrasConfig: Cerebras-specific configuration for a model.
              - …
            - AIGatewayTargetCohereConfig: Cohere-specific configuration for a model.
              - …
            - AIGatewayTargetDashscopeConfig: Alibaba DashScope-specific configuration for a model.
              - …
            - AIGatewayTargetDatabricksConfig: Databricks-specific configuration for a model.
              - …
            - AIGatewayTargetDeepseekConfig: Deepseek-specific configuration for a model.
              - …
            - AIGatewayTargetGeminiConfig: Google Gemini-specific configuration for a model.
              - …
            - AIGatewayTargetHuggingfaceConfig: Hugging Face-specific configuration for a model.
              - …
            - AIGatewayTargetKimiConfig: Kimi (Moonshot AI)-specific configuration for a model.
              - …
            - AIGatewayTargetLlama2Config: Llama2-specific configuration for a model.
              - …
            - AIGatewayTargetMistralConfig: Mistral-specific configuration for a model.
              - …
            - AIGatewayTargetOllamaConfig: Ollama-specific configuration for a model.
              - …
            - AIGatewayTargetOpenaiConfig: Openai-specific configuration for a model.
              - …
            - AIGatewayTargetVercelConfig: Vercel AI Gateway-specific configuration for a model.
              - …
            - AIGatewayTargetVllmConfig: Vllm-specific configuration for a model.
              - …
            - AIGatewayTargetXaiConfig: Xai-specific configuration for a model.
              - …
            - AIGatewayTargetSagemakerConfig: AWS SageMaker-specific configuration for a model.
              - …
            - AIGatewayTargetTypesafeConfig: Typesafe-specific configuration for a model. **Requires a minimum runtime version of `2.2`**.
              - …
        - `policies` string[]: List of policy references.
        - `labels` PublicLabels: Public labels store information about an entity that can be used for filtering a list of objects. Public labels are intended to store **PUBLIC** metadata. Keys must be of length 1-63 characters, and cannot start with "kong", "konnect", "mesh", "kic", or "_".
        - `managed_by` ManagedBy: Stores information about what manages this entity, such as the tool or system responsible for its lifecycle (for example, `terraform`). Keys must be 1–63 characters long, start and end with an alphanumeric character, and may contain dashes (`-`), underscores (`_`), and dots (`.`). Values must be 1–127 characters long and must not be empty. Keys and values must not contain spaces or control characters.
        - `datastores` AIGatewayDatastoreRef[]: Names of the Datastores this model references.
          - `name` string, required: An immutable user-defined identifier for this resource.
        - `type` 'api', required
        - `config` AIGatewayModelAPIConfig, required: Routing, logging, and load balancing configuration for the model.
          - `route` AIGatewayModelRouteConfig, required: Configuration for an AI Gateway route.
            - `headers` object: One or more lists of values indexed by header name that will cause this route to match if present in the request. The `Host` header cannot be used with this attribute: hosts should be specified using the `hosts` attribute. When `headers` contains only one value and that value starts with the special prefix `~*`, the value is interpreted as a regular expression.
            - `hosts` string[]: A list of domain names that match this route. Note that the hosts value is case sensitive.
            - `https_redirect_status_code` integer: The status code Kong responds with when all properties of a route match except the protocol i.e. if the protocol of the request is `HTTP` instead of `HTTPS`. `Location` header is injected by Kong if the field is set to 301, 302, 307 or 308. Note: This config applies only if the route is configured to only accept the `https` protocol.
            - `methods` string[]: A list of HTTP methods that match this route.
            - `paths` string[]: A list of paths that match this route.
            - `preserve_host` boolean: When matching a route via one of the `hosts` domain names, use the request `Host` header in the upstream request headers. If set to `false`, the upstream `Host` header will be that of the service's `host`.
            - `protocols` string[]: An array of the protocols this route should allow. See the [route Object](#route-object) section for a list of accepted protocols. When set to only `https`, HTTP requests are answered with an upgrade error. When set to only `http`, HTTPS requests are answered with an error.
            - `regex_priority` integer: A number used to choose which route resolves a given request when several routes match it using regexes simultaneously. When two routes match the path and have the same `regex_priority`, the older one (lowest `created_at`) is used. Note that the priority for non-regex routes is different (longer non-regex routes are matched before shorter ones).
            - `request_buffering` boolean: Whether to enable request body buffering or not. With HTTP 1.1, it may make sense to turn this off on services that receive data with chunked transfer encoding.
            - `response_buffering` boolean: Whether to enable response body buffering or not. With HTTP 1.1, it may make sense to turn this off on services that send data with chunked transfer encoding.
            - `strip_path` boolean: When matching a route via one of the `paths`, strip the matching prefix from the upstream request URL.
            - `tags` string[]: An optional set of strings associated with the route for grouping and filtering.
            - `model` AIGatewayModelSelectorConfig: Configuration for overriding routing to this model using a selector. When no selector location is set, the format default selector is used. When values are not set, the model name is used as the selector value.
              - …
          - `logging` AIGatewayLoggingConfig: Configuration for AI Gateway logging.
            - `payloads` boolean
          - `response_streaming` 'allow' | 'always' | 'deny'
          - `max_request_body_size` integer: Maximum size of request body to parse. Set to 0 for unlimited.
          - `balancer` union: Configuration for a model's load balancer when multiple target models are configured.
            - AIGatewayModelBalancerConsistentHashingConfig
              - …
            - AIGatewayModelBalancerLeastConnectionsConfig
              - …
            - AIGatewayModelBalancerLowestLatencyConfig
              - …
            - AIGatewayModelBalancerLowestUsageConfig
              - …
            - AIGatewayModelBalancerPriorityConfig
              - …
            - AIGatewayModelBalancerRoundRobinConfig
              - …
            - AIGatewayModelBalancerSemanticConfig
              - …
          - `proxy` AIGatewayProxyConfig: HTTP/HTTPS proxy configuration for outbound requests to the upstream AI provider.
            - `http_proxy` object: HTTP proxy server to route plaintext outbound requests through.
              - …
            - `https_proxy` object: HTTPS proxy server to route TLS outbound requests through.
              - …
            - `proxy_scheme` 'http': The proxy scheme to use when connecting to the proxy server.
            - `auth` object: Credentials used to authenticate to the proxy server.
              - …
            - `no_proxy` string: Comma-separated list of hosts that should not be proxied.
        - `capabilities` string[], required: The capabilities this model supports. Set to a specific capability to restrict the model to that capability only, or to `*` to allow all capabilities that the targets support at the gateway's min_runtime_version. A capability that needs a newer min_runtime_version is left out of `*`.
      - object: Configuration for proxying synchronous requests/responses to/from an AI Gateway model using generative APIs.
        - `id` string, uuid, required: Contains a unique identifier used for this resource.
        - `created_at` string, date-time, required: An ISO-8601 timestamp representation of entity creation date.
        - `updated_at` string, date-time, required: An ISO-8601 timestamp representation of entity update date.
        - `display_name` string, required: The display name for this model instance.
        - `name` string, required: Identifier for an AI Gateway entity. In some cases, this may be the entity name or ID.
        - `enabled` boolean: Whether the model is enabled.
        - `access` AIGatewayModelAccess: Access control configuration for a model.
          - `acls` union: Access control rules. Configure exactly one of `allow`, `deny`, `allow_when`, or `deny_when`.
            - AIGatewayAllowACL
              - …
            - AIGatewayDenyACL
              - …
            - AIGatewayAllowWhenACL
              - …
            - AIGatewayDenyWhenACL
              - …
          - `identity_providers` AIGatewayIdentityProviderReference[]: List of identity providers for granting access to the model. At most 1 identity provider of each identity provider type can be referenced. Deprecated: use `auth_strategies` instead. The two are mutually exclusive.
          - `auth_strategies` AIGatewayAuthStrategyReference[]: List of auth strategies for granting access to the model. At most 1 auth strategy of each auth strategy type can be referenced.
        - `formats` AIGatewayModelFormat[], required: List of request/response formats supported by this model.
          - `type` '*' | 'anthropic' | 'bedrock' | 'cohere' | 'gemini' | 'huggingface' | 'openai' | 'passthrough' | 'typesafe': The request and response format this model supports. Set to a specific format to restrict the model to that format only, or to `*` to allow all formats that the targets support at the gateway's min_runtime_version. A format that needs a newer min_runtime_version is left out of `*`. The `*` format never includes passthrough. **`passthrough` requires a minimum runtime version of `2.2`**. **`typesafe` requires a minimum runtime version of `2.2`**.
        - `targets` AIGatewayTarget[], required: One or more backend models that this model entry routes to.
          - `name` string, required: The name of the model defined in the upstream provider that will be executed.
          - `weight` integer: The weight this target gets within the upstream load balancer
          - `semantic_description` string: The semantic description of the target, required if using semantic load balancing. Specially, setting this to 'CATCHALL' will indicate such target to be used when no other targets match the semantic threshold.
          - `allow_auth_override` boolean: When enabled, request-level auth parameters (such as API keys or bearer tokens) will override the static values defined for the provider.
          - `provider` string, required: Reference to a model provider instance by name.
          - `config` union, required: Configuration for a target model.
            - AIGatewayTargetAnthropicConfig: Anthropic-specific configuration for a model.
              - …
            - AIGatewayTargetAzureConfig: Azure-specific configuration for a model.
              - …
            - AIGatewayTargetBasetenConfig: Baseten-specific configuration for a model. Requests are sent to Baseten Model APIs at `https://inference.baseten.co/v1` by default, and the target `name` is the Baseten model slug (for example, `deepseek-ai/DeepSeek-V4-Pro-0813`). To route to a dedicated Baseten deployment, set `upstream_url` to the deployment's full OpenAI-compatible endpoint (for example, `https://model-{model_id}.api.baseten.co/environments/{environment}/sync/v1/chat/completions`) or use passthrough format. **Requires a minimum runtime version of `2.3`**.
              - …
            - AIGatewayTargetBedrockConfig: AWS Bedrock-specific configuration for a model.
              - …
            - AIGatewayTargetCerebrasConfig: Cerebras-specific configuration for a model.
              - …
            - AIGatewayTargetCohereConfig: Cohere-specific configuration for a model.
              - …
            - AIGatewayTargetDashscopeConfig: Alibaba DashScope-specific configuration for a model.
              - …
            - AIGatewayTargetDatabricksConfig: Databricks-specific configuration for a model.
              - …
            - AIGatewayTargetDeepseekConfig: Deepseek-specific configuration for a model.
              - …
            - AIGatewayTargetGeminiConfig: Google Gemini-specific configuration for a model.
              - …
            - AIGatewayTargetHuggingfaceConfig: Hugging Face-specific configuration for a model.
              - …
            - AIGatewayTargetKimiConfig: Kimi (Moonshot AI)-specific configuration for a model.
              - …
            - AIGatewayTargetLlama2Config: Llama2-specific configuration for a model.
              - …
            - AIGatewayTargetMistralConfig: Mistral-specific configuration for a model.
              - …
            - AIGatewayTargetOllamaConfig: Ollama-specific configuration for a model.
              - …
            - AIGatewayTargetOpenaiConfig: Openai-specific configuration for a model.
              - …
            - AIGatewayTargetVercelConfig: Vercel AI Gateway-specific configuration for a model.
              - …
            - AIGatewayTargetVllmConfig: Vllm-specific configuration for a model.
              - …
            - AIGatewayTargetXaiConfig: Xai-specific configuration for a model.
              - …
            - AIGatewayTargetSagemakerConfig: AWS SageMaker-specific configuration for a model.
              - …
            - AIGatewayTargetTypesafeConfig: Typesafe-specific configuration for a model. **Requires a minimum runtime version of `2.2`**.
              - …
        - `policies` string[]: List of policy references.
        - `labels` PublicLabels: Public labels store information about an entity that can be used for filtering a list of objects. Public labels are intended to store **PUBLIC** metadata. Keys must be of length 1-63 characters, and cannot start with "kong", "konnect", "mesh", "kic", or "_".
        - `managed_by` ManagedBy: Stores information about what manages this entity, such as the tool or system responsible for its lifecycle (for example, `terraform`). Keys must be 1–63 characters long, start and end with an alphanumeric character, and may contain dashes (`-`), underscores (`_`), and dots (`.`). Values must be 1–127 characters long and must not be empty. Keys and values must not contain spaces or control characters.
        - `datastores` AIGatewayDatastoreRef[]: Names of the Datastores this model references.
          - `name` string, required: An immutable user-defined identifier for this resource.
        - `type` 'model', required
        - `config` AIGatewayModelModelConfig, required: Routing, logging, and load balancing configuration for the model.
          - `route` AIGatewayModelRouteConfig, required: Configuration for an AI Gateway route.
            - `headers` object: One or more lists of values indexed by header name that will cause this route to match if present in the request. The `Host` header cannot be used with this attribute: hosts should be specified using the `hosts` attribute. When `headers` contains only one value and that value starts with the special prefix `~*`, the value is interpreted as a regular expression.
            - `hosts` string[]: A list of domain names that match this route. Note that the hosts value is case sensitive.
            - `https_redirect_status_code` integer: The status code Kong responds with when all properties of a route match except the protocol i.e. if the protocol of the request is `HTTP` instead of `HTTPS`. `Location` header is injected by Kong if the field is set to 301, 302, 307 or 308. Note: This config applies only if the route is configured to only accept the `https` protocol.
            - `methods` string[]: A list of HTTP methods that match this route.
            - `paths` string[]: A list of paths that match this route.
            - `preserve_host` boolean: When matching a route via one of the `hosts` domain names, use the request `Host` header in the upstream request headers. If set to `false`, the upstream `Host` header will be that of the service's `host`.
            - `protocols` string[]: An array of the protocols this route should allow. See the [route Object](#route-object) section for a list of accepted protocols. When set to only `https`, HTTP requests are answered with an upgrade error. When set to only `http`, HTTPS requests are answered with an error.
            - `regex_priority` integer: A number used to choose which route resolves a given request when several routes match it using regexes simultaneously. When two routes match the path and have the same `regex_priority`, the older one (lowest `created_at`) is used. Note that the priority for non-regex routes is different (longer non-regex routes are matched before shorter ones).
            - `request_buffering` boolean: Whether to enable request body buffering or not. With HTTP 1.1, it may make sense to turn this off on services that receive data with chunked transfer encoding.
            - `response_buffering` boolean: Whether to enable response body buffering or not. With HTTP 1.1, it may make sense to turn this off on services that send data with chunked transfer encoding.
            - `strip_path` boolean: When matching a route via one of the `paths`, strip the matching prefix from the upstream request URL.
            - `tags` string[]: An optional set of strings associated with the route for grouping and filtering.
            - `model` AIGatewayModelSelectorConfig: Configuration for overriding routing to this model using a selector. When no selector location is set, the format default selector is used. When values are not set, the model name is used as the selector value.
              - …
          - `logging` AIGatewayLoggingConfig: Configuration for AI Gateway logging.
            - `payloads` boolean
          - `response_streaming` 'allow' | 'always' | 'deny'
          - `max_request_body_size` integer: Maximum size of request body to parse. Set to 0 for unlimited.
          - `model` object
            - `name_header` boolean: Display the model name selected in the X-Kong-LLM-Model response header
          - `balancer` union: Configuration for a model's load balancer when multiple target models are configured.
            - AIGatewayModelBalancerConsistentHashingConfig
              - …
            - AIGatewayModelBalancerLeastConnectionsConfig
              - …
            - AIGatewayModelBalancerLowestLatencyConfig
              - …
            - AIGatewayModelBalancerLowestUsageConfig
              - …
            - AIGatewayModelBalancerPriorityConfig
              - …
            - AIGatewayModelBalancerRoundRobinConfig
              - …
            - AIGatewayModelBalancerSemanticConfig
              - …
          - `proxy` AIGatewayProxyConfig: HTTP/HTTPS proxy configuration for outbound requests to the upstream AI provider.
            - `http_proxy` object: HTTP proxy server to route plaintext outbound requests through.
              - …
            - `https_proxy` object: HTTPS proxy server to route TLS outbound requests through.
              - …
            - `proxy_scheme` 'http': The proxy scheme to use when connecting to the proxy server.
            - `auth` object: Credentials used to authenticate to the proxy server.
              - …
            - `no_proxy` string: Comma-separated list of hosts that should not be proxied.
        - `capabilities` string[], required: The capabilities this model supports. Set to a specific capability to restrict the model to that capability only, or to `*` to allow all capabilities that the targets support at the gateway's min_runtime_version. A capability that needs a newer min_runtime_version is left out of `*`.
  - `mcp_servers` AIGatewayMCPServer[], required
    - union
      - object
        - `id` string, uuid, required: Contains a unique identifier used for this resource.
        - `created_at` string, date-time, required: An ISO-8601 timestamp representation of entity creation date.
        - `updated_at` string, date-time, required: An ISO-8601 timestamp representation of entity update date.
        - `type` 'conversion-only', required
        - `config` AIGatewayMCPServerWithUpstreamNoProxyConfigNoServerConfig, required: Routing, logging, and request body size limits for the MCP Server.
          - `route` union
            - object: Configuration for an AI Gateway route.
              - …
            - object: Configuration for an AI Gateway route.
              - …
            - object: Configuration for an AI Gateway route.
              - …
            - object: Configuration for an AI Gateway route.
              - …
          - `logging` object: Configuration for AI Gateway logging.
            - `payloads` boolean
            - `audits` boolean
          - `max_request_body_size` integer: Maximum size of request body to parse. Set to 0 for unlimited.
          - `url` string, uri, required: Helper field to set protocol, host, port and path of the upstream service using a URL. This is the same as a Kong Gateway Service URL: ${scheme}://${host}:${port}/${path}
          - `upstream` AIGatewayUpstreamConfig: Configuration applied when proxying to the upstream service, including authentication.
            - `auth` AIGatewayUpstreamAuthAWS: AWS IAM (SigV4) authentication for the upstream service.
              - …
          - `certificates` AIGatewayCertificateReference[]: Client certificates used for mutual TLS (mTLS) authentication with the upstream MCP server, referenced by name. Each entry must be the `name` of an existing AI Certificate on the same AI Gateway. The certificates are applied to the Kong Gateway Service generated for this MCP Server, which proxies requests to the upstream MCP server. The generated Service presents a single client certificate (`client_certificate`) during the TLS handshake with the upstream, so at most one certificate may be configured. If unset, no client certificate is presented to the upstream.
          - `ca_certificates` AIGatewayCACertificateReference[]: CA certificates used to verify the TLS certificate of the upstream MCP server, referenced by name. Each entry must be the `name` of an existing AI CA Certificate on the same AI Gateway. The CA certificates build the trust store (`ca_certificates`) of the Kong Gateway Service generated for this MCP Server, which proxies requests to the upstream MCP server. If unset, the default trust store is used to verify the upstream's TLS certificate.
        - `tools` AIGatewayMCPConversionTool[], required: List of tools exposed by this MCP Server.
          - `access` AIGatewayMCPToolAccess: Access-control rules for a tool.
            - `acls` AIGatewayMCPACLs: Access control rules for MCP resources. Configure `allow`, `deny`, or both.
              - …
          - `annotations` AIGatewayMCPToolAnnotations
            - `destructive_hint` boolean: If true, the tool may perform destructive updates
            - `idempotent_hint` boolean: If true, repeated calls with same args have no additional effect
            - `open_world_hint` boolean: If true, tool interacts with external entities
            - `read_only_hint` boolean: If true, the tool does not modify its environment
            - `title` string: Human-readable title for the tool
          - `description` string, required: A description of what the tool does.
          - `name` string, required: The MCP tool name. In upstream-server mode, it also matches the remote MCP Server tool whose metadata this entry overrides.
          - `headers` AIGatewayMCPToolHeaders: The headers of the exported API. By default, Kong will extract the headers from API configuration. If the configured headers are not exactly matched, this field is required.
          - `host` string: The host used when forwarding the request to the upstream API. By default, Kong will extract the host from API configuration. If the configured host is wildcard, this field is required.
          - `method` 'DELETE' | 'GET' | 'PATCH' | 'POST' | 'PUT', required: The HTTP method used when forwarding the request to the upstream API.
          - `path` string: The path of the exported API. Always treated as relative to the path component of `config.url` and simply concatenated onto it — a leading `/` has no special "absolute path" meaning. If this tool's `host` or `scheme` overrides the source's URL, `path` is instead relative to the root of that overridden host, since there is no URL path from a different host to append to. By default, Kong will extract the path from API configuration.
          - `query` AIGatewayMCPToolQuery: The query arguments of the exported API. If the generated query arguments are not exactly matched, this field is required.
          - `request_body` AIGatewayMCPToolRequestBody: The API requestBody specification defined in OpenAPI JSON format. For example, '{"content":{"application/x-www-form-urlencoded":{"schema":{"type":"object","properties":{"color":{"type":"array","items":{"type":"string"}}}}}}}'. See https://swagger.io/docs/specification/v3_0/describing-request-body/describing-request-body/ for more details. Note that `$ref` is not supported.
          - `responses` AIGatewayMCPToolResponses: The API responses specification defined in OpenAPI JSON format. This specification will be used to validate the upstream response and map it back to the structuredOutput. For example, '{"200":{"content":{"application/json":{"schema":{"type":"object","properties":{"result":{"type":"string"}}}}}}}}'. See https://swagger.io/docs/specification/v3_0/describing-responses/ for more details. Only one non-error (status code < 400) response is supported. Note that `$ref` is not supported.
          - `scheme` 'http' | 'https': The scheme of the exported API. By default, Kong will extract the scheme from API configuration. If the configured scheme is not expected, this field can be used to override it.
          - `parameters` AIGatewayMCPToolParameter[]
            - `name` string, required: The name of the parameter.
            - `in` 'query' | 'path' | 'header' | 'body', required: The location of the parameter in the request.
            - `description` string: A description of the parameter.
            - `required` boolean: Whether this parameter is required.
            - `schema` object: JSON Schema definition for the parameter value. See https://swagger.io/docs/specification/v3_0/describing-parameters/#schema-vs-content for more details.
        - `display_name` string, required: The display name for the MCP Server.
        - `name` string, required: Identifier for an AI Gateway entity. In some cases, this may be the entity name or ID.
        - `enabled` boolean: Whether the MCP Server is enabled.
        - `policies` string[]: List of policy references.
        - `labels` PublicLabels: Public labels store information about an entity that can be used for filtering a list of objects. Public labels are intended to store **PUBLIC** metadata. Keys must be of length 1-63 characters, and cannot start with "kong", "konnect", "mesh", "kic", or "_".
        - `managed_by` ManagedBy: Stores information about what manages this entity, such as the tool or system responsible for its lifecycle (for example, `terraform`). Keys must be 1–63 characters long, start and end with an alphanumeric character, and may contain dashes (`-`), underscores (`_`), and dots (`.`). Values must be 1–127 characters long and must not be empty. Keys and values must not contain spaces or control characters.
        - `datastores` AIGatewayDatastoreRef[]: The Token Vault Datastore holding this entity's credential storage: the directory credentials are written to, and the encryption keys that seal them. Required when `token_vault` is set (on an auth strategy, when its credential manager is enabled), and the referenced Datastore must have type `token-vault`. Only one Datastore can be referenced. **Requires a minimum runtime version of `2.3`**.
          - `name` string, required: An immutable user-defined identifier for this resource.
        - `token_vault` AIGatewayTokenVault: Resolves an upstream credential per request via Kong's Token Vault instead of sending a static credential. The directory credentials live in, and the keys that encrypt them, come from the `token-vault` Datastore this entity references, and are returned read-only in `storage`. Callers must enroll with the configured Token Vault provider before the credential can be resolved. Until enrollment completes, an MCP Server serves virtual `authenticate` and `check_authentication_status` tools that guide the caller through the enrollment flow. **Requires a minimum runtime version of `2.3`**.
          - `provider` string, required: Name of the upstream credential provider registered in the Token Vault directory.
          - `subject_token_type` 'access_token' | 'api_key': The kind of bearer token the caller presents, which the Token Vault exchanges for the upstream credential. Omit it unless the request shape makes the token type ambiguous: it is otherwise resolved from the request, and failing that from the token's shape (a JWS is an `access_token`, anything else an `api_key`).
          - `credential` AIGatewayTokenVaultCredential: Where the caller's subject token is read from, and where the credential the Token Vault returns is placed on the upstream request. Both placements default to `Authorization: Bearer <credential>` in a header, though an `outbound` placement left unset honours whatever the provider advertises instead. Set them for providers that take the credential elsewhere, such as an `x-api-key` header carrying the raw credential.
            - `inbound` AIGatewayTokenVaultCredentialPlacement
              - …
            - `outbound` AIGatewayTokenVaultCredentialPlacement
              - …
          - `cache` object: Optional best-effort cache for exchanged credentials, shared across the cluster. Losing it only costs a re-exchange. Credentials are encrypted with the keys held by the Token Vault Datastore before being cached.
            - `redis` AIGatewayRedisCloudConfiguration: Config for connecting to a Cloud Provider's Redis instance.
              - …
          - `storage` TokenVaultDatastoreConfig
            - `directory` string, required: The name of the directory in which tokens are stored.
            - `current_version` string, required: The version of the encryption key used to encrypt new tokens.
            - `keys` TokenVaultDatastoreKey[], required: The encryption keys used to protect stored tokens. Keep previous versions in the list so existing tokens can still be decrypted.
              - …
      - object
        - `id` string, uuid, required: Contains a unique identifier used for this resource.
        - `created_at` string, date-time, required: An ISO-8601 timestamp representation of entity creation date.
        - `updated_at` string, date-time, required: An ISO-8601 timestamp representation of entity update date.
        - `type` 'conversion-listener', required
        - `config` object, required: Server-side configuration specific to modes where Kong answers as the MCP server.
          - `route` union, required
            - object: Configuration for an AI Gateway route.
              - …
            - object: Configuration for an AI Gateway route.
              - …
            - object: Configuration for an AI Gateway route.
              - …
            - object: Configuration for an AI Gateway route.
              - …
          - `logging` object: Configuration for AI Gateway logging.
            - `payloads` boolean
            - `audits` boolean
          - `max_request_body_size` integer: Maximum size of request body to parse. Set to 0 for unlimited.
          - `server` AIGatewayMCPServerServerConfigBase: Server-side configuration for the MCP Server.
            - `forward_client_headers` boolean: Whether to forward the client request headers to the upstream server when calling the tools.
            - `session` object: Enable managed session when Kong responds as MCP server in listener, conversion-listener, or upstream-server modes. This doesn't affect the passthrough-listener mode as the state in that mode is maintained by the upstream MCP servers.
              - …
            - `timeout` integer: The timeout for calling the tools in milliseconds.
          - `url` string, uri, required: Helper field to set protocol, host, port and path of the upstream service using a URL. This is the same as a Kong Gateway Service URL: ${scheme}://${host}:${port}/${path}
          - `upstream` AIGatewayUpstreamConfig: Configuration applied when proxying to the upstream service, including authentication.
            - `auth` AIGatewayUpstreamAuthAWS: AWS IAM (SigV4) authentication for the upstream service.
              - …
          - `certificates` AIGatewayCertificateReference[]: Client certificates used for mutual TLS (mTLS) authentication with the upstream MCP server, referenced by name. Each entry must be the `name` of an existing AI Certificate on the same AI Gateway. The certificates are applied to the Kong Gateway Service generated for this MCP Server, which proxies requests to the upstream MCP server. The generated Service presents a single client certificate (`client_certificate`) during the TLS handshake with the upstream, so at most one certificate may be configured. If unset, no client certificate is presented to the upstream.
          - `ca_certificates` AIGatewayCACertificateReference[]: CA certificates used to verify the TLS certificate of the upstream MCP server, referenced by name. Each entry must be the `name` of an existing AI CA Certificate on the same AI Gateway. The CA certificates build the trust store (`ca_certificates`) of the Kong Gateway Service generated for this MCP Server, which proxies requests to the upstream MCP server. If unset, the default trust store is used to verify the upstream's TLS certificate.
          - `allowed_versions` string[]: The MCP protocol revisions this server accepts. Leave unset to accept every revision Kong implements, which is the default. When set, `server/discover` advertises exactly this list and a request declaring anything else is rejected. Listing only per-request revisions refuses handshake clients. **Requires a minimum runtime version of `2.1`**.
          - `cache` object: Cache hints Kong emits on the cacheable operations it serves. Only clients on a protocol revision that defines them receive them. **Requires a minimum runtime version of `2.1`**.
            - `tools_list` AIGatewayMCPServerCacheHint: A cache hint Kong emits on a cacheable operation it serves.
              - …
            - `discover` AIGatewayMCPServerCacheHint: A cache hint Kong emits on a cacheable operation it serves.
              - …
        - `tools` AIGatewayMCPConversionTool[], required: List of tools exposed by this MCP Server. Each tool's `path`, `method`, and `host` describe the backend HTTP operation on the upstream selected by `config.url` — they do not need to match the public MCP Route configured in `config.route`.
          - `access` AIGatewayMCPToolAccess: Access-control rules for a tool.
            - `acls` AIGatewayMCPACLs: Access control rules for MCP resources. Configure `allow`, `deny`, or both.
              - …
          - `annotations` AIGatewayMCPToolAnnotations
            - `destructive_hint` boolean: If true, the tool may perform destructive updates
            - `idempotent_hint` boolean: If true, repeated calls with same args have no additional effect
            - `open_world_hint` boolean: If true, tool interacts with external entities
            - `read_only_hint` boolean: If true, the tool does not modify its environment
            - `title` string: Human-readable title for the tool
          - `description` string, required: A description of what the tool does.
          - `name` string, required: The MCP tool name. In upstream-server mode, it also matches the remote MCP Server tool whose metadata this entry overrides.
          - `headers` AIGatewayMCPToolHeaders: The headers of the exported API. By default, Kong will extract the headers from API configuration. If the configured headers are not exactly matched, this field is required.
          - `host` string: The host used when forwarding the request to the upstream API. By default, Kong will extract the host from API configuration. If the configured host is wildcard, this field is required.
          - `method` 'DELETE' | 'GET' | 'PATCH' | 'POST' | 'PUT', required: The HTTP method used when forwarding the request to the upstream API.
          - `path` string: The path of the exported API. Always treated as relative to the path component of `config.url` and simply concatenated onto it — a leading `/` has no special "absolute path" meaning. If this tool's `host` or `scheme` overrides the source's URL, `path` is instead relative to the root of that overridden host, since there is no URL path from a different host to append to. By default, Kong will extract the path from API configuration.
          - `query` AIGatewayMCPToolQuery: The query arguments of the exported API. If the generated query arguments are not exactly matched, this field is required.
          - `request_body` AIGatewayMCPToolRequestBody: The API requestBody specification defined in OpenAPI JSON format. For example, '{"content":{"application/x-www-form-urlencoded":{"schema":{"type":"object","properties":{"color":{"type":"array","items":{"type":"string"}}}}}}}'. See https://swagger.io/docs/specification/v3_0/describing-request-body/describing-request-body/ for more details. Note that `$ref` is not supported.
          - `responses` AIGatewayMCPToolResponses: The API responses specification defined in OpenAPI JSON format. This specification will be used to validate the upstream response and map it back to the structuredOutput. For example, '{"200":{"content":{"application/json":{"schema":{"type":"object","properties":{"result":{"type":"string"}}}}}}}}'. See https://swagger.io/docs/specification/v3_0/describing-responses/ for more details. Only one non-error (status code < 400) response is supported. Note that `$ref` is not supported.
          - `scheme` 'http' | 'https': The scheme of the exported API. By default, Kong will extract the scheme from API configuration. If the configured scheme is not expected, this field can be used to override it.
          - `parameters` AIGatewayMCPToolParameter[]
            - `name` string, required: The name of the parameter.
            - `in` 'query' | 'path' | 'header' | 'body', required: The location of the parameter in the request.
            - `description` string: A description of the parameter.
            - `required` boolean: Whether this parameter is required.
            - `schema` object: JSON Schema definition for the parameter value. See https://swagger.io/docs/specification/v3_0/describing-parameters/#schema-vs-content for more details.
        - `access` union
          - AIGatewayMCPServerListenerConsumer: Auth strategy and OAuth 2.0 Protected Resource Metadata configuration for granting access to an MCP server.
            - `acl_attribute_type` 'consumer', required: The type of attributes that ACL is evaluated with.
            - `acls` union: Access control rules for MCP resources. Configure exactly one of `allow`, `deny`, `allow_when`, or `deny_when`.
              - …
            - `default_tool_acls` AIGatewayMCPACLs: Access control rules for MCP resources. Configure `allow`, `deny`, or both.
              - …
            - `identity_providers` AIGatewayIdentityProviderReference[]: List of identity providers for granting access to the MCP server. At most 1 identity provider of each identity provider type can be referenced. Deprecated: use `auth_strategies` instead. The two are mutually exclusive.
            - `auth_strategies` AIGatewayAuthStrategyReference[]: List of auth strategies for granting access to the MCP server. At most 1 auth strategy of each auth strategy type can be referenced.
            - `metadata` AIGatewayMCPServerProtectedResourceMetadata: OAuth 2.0 Protected Resource Metadata (RFC 9728) advertised for this MCP server, allowing clients to discover the authorization servers that protect it.
              - …
          - AIGatewayMCPServerListenerOauth: Auth strategy and OAuth 2.0 Protected Resource Metadata configuration for granting access to an MCP server.
            - `acl_attribute_type` 'oauth_access_token', required: The type of attributes that ACL is evaluated with.
            - `access_token_claim_field` string, required: The claim in the OAuth2 access token to use as the subject for ACL evaluation when `acl_attribute_type` is set to `oauth_access_token`. Nested claim can be fetched by using a jq filter starts with dot, e.g., “.user.email”: https://jqlang.org/manual/#object-identifier-index
            - `acls` union: Access control rules for MCP resources. Configure exactly one of `allow`, `deny`, `allow_when`, or `deny_when`.
              - …
            - `default_tool_acls` AIGatewayMCPACLs: Access control rules for MCP resources. Configure `allow`, `deny`, or both.
              - …
            - `identity_providers` AIGatewayIdentityProviderReference[]: List of identity providers for granting access to the MCP server. At most 1 identity provider of each identity provider type can be referenced. Deprecated: use `auth_strategies` instead. The two are mutually exclusive.
            - `auth_strategies` AIGatewayAuthStrategyReference[]: List of auth strategies for granting access to the MCP server. At most 1 auth strategy of each auth strategy type can be referenced.
            - `metadata` AIGatewayMCPServerProtectedResourceMetadata: OAuth 2.0 Protected Resource Metadata (RFC 9728) advertised for this MCP server, allowing clients to discover the authorization servers that protect it.
              - …
        - `rate_limiting` AIGatewayMCPServerRateLimiting: Rate limits applied to requests to this MCP Server, matched by MCP method or tool name. Each AI Gateway node counts and enforces the limits independently. **Requires a minimum runtime version of `2.3`**.
          - `limits` AIGatewayMCPServerRateLimit[], required: The rate limits to enforce. Each entry targets exactly one of `method` or `tool`. Two entries cannot share the same target and `window_size`.
            - `method` string: The MCP method this limit applies to.
            - `tool` string: The name of the tool this limit applies to.
            - `limit` integer, required: The number of requests allowed per window.
            - `window_size` 1 | 60 | 3600 | 86400, required: The window size in seconds, one of a second, a minute, an hour, or a day.
        - `display_name` string, required: The display name for the MCP Server.
        - `name` string, required: Identifier for an AI Gateway entity. In some cases, this may be the entity name or ID.
        - `enabled` boolean: Whether the MCP Server is enabled.
        - `policies` string[]: List of policy references.
        - `labels` PublicLabels: Public labels store information about an entity that can be used for filtering a list of objects. Public labels are intended to store **PUBLIC** metadata. Keys must be of length 1-63 characters, and cannot start with "kong", "konnect", "mesh", "kic", or "_".
        - `managed_by` ManagedBy: Stores information about what manages this entity, such as the tool or system responsible for its lifecycle (for example, `terraform`). Keys must be 1–63 characters long, start and end with an alphanumeric character, and may contain dashes (`-`), underscores (`_`), and dots (`.`). Values must be 1–127 characters long and must not be empty. Keys and values must not contain spaces or control characters.
        - `datastores` AIGatewayDatastoreRef[]: The Token Vault Datastore holding this entity's credential storage: the directory credentials are written to, and the encryption keys that seal them. Required when `token_vault` is set (on an auth strategy, when its credential manager is enabled), and the referenced Datastore must have type `token-vault`. Only one Datastore can be referenced. **Requires a minimum runtime version of `2.3`**.
          - `name` string, required: An immutable user-defined identifier for this resource.
        - `token_vault` AIGatewayTokenVault: Resolves an upstream credential per request via Kong's Token Vault instead of sending a static credential. The directory credentials live in, and the keys that encrypt them, come from the `token-vault` Datastore this entity references, and are returned read-only in `storage`. Callers must enroll with the configured Token Vault provider before the credential can be resolved. Until enrollment completes, an MCP Server serves virtual `authenticate` and `check_authentication_status` tools that guide the caller through the enrollment flow. **Requires a minimum runtime version of `2.3`**.
          - `provider` string, required: Name of the upstream credential provider registered in the Token Vault directory.
          - `subject_token_type` 'access_token' | 'api_key': The kind of bearer token the caller presents, which the Token Vault exchanges for the upstream credential. Omit it unless the request shape makes the token type ambiguous: it is otherwise resolved from the request, and failing that from the token's shape (a JWS is an `access_token`, anything else an `api_key`).
          - `credential` AIGatewayTokenVaultCredential: Where the caller's subject token is read from, and where the credential the Token Vault returns is placed on the upstream request. Both placements default to `Authorization: Bearer <credential>` in a header, though an `outbound` placement left unset honours whatever the provider advertises instead. Set them for providers that take the credential elsewhere, such as an `x-api-key` header carrying the raw credential.
            - `inbound` AIGatewayTokenVaultCredentialPlacement
              - …
            - `outbound` AIGatewayTokenVaultCredentialPlacement
              - …
          - `cache` object: Optional best-effort cache for exchanged credentials, shared across the cluster. Losing it only costs a re-exchange. Credentials are encrypted with the keys held by the Token Vault Datastore before being cached.
            - `redis` AIGatewayRedisCloudConfiguration: Config for connecting to a Cloud Provider's Redis instance.
              - …
          - `storage` TokenVaultDatastoreConfig
            - `directory` string, required: The name of the directory in which tokens are stored.
            - `current_version` string, required: The version of the encryption key used to encrypt new tokens.
            - `keys` TokenVaultDatastoreKey[], required: The encryption keys used to protect stored tokens. Keep previous versions in the list so existing tokens can still be decrypted.
              - …
      - object
        - `id` string, uuid, required: Contains a unique identifier used for this resource.
        - `created_at` string, date-time, required: An ISO-8601 timestamp representation of entity creation date.
        - `updated_at` string, date-time, required: An ISO-8601 timestamp representation of entity update date.
        - `type` 'listener', required
        - `config` object, required: Server-side configuration specific to modes where Kong answers as the MCP server.
          - `route` union, required
            - object: Configuration for an AI Gateway route.
              - …
            - object: Configuration for an AI Gateway route.
              - …
            - object: Configuration for an AI Gateway route.
              - …
            - object: Configuration for an AI Gateway route.
              - …
          - `logging` object: Configuration for AI Gateway logging.
            - `payloads` boolean
            - `audits` boolean
          - `max_request_body_size` integer: Maximum size of request body to parse. Set to 0 for unlimited.
          - `server` AIGatewayMCPServerServerConfigBase: Server-side configuration for the MCP Server.
            - `forward_client_headers` boolean: Whether to forward the client request headers to the upstream server when calling the tools.
            - `session` object: Enable managed session when Kong responds as MCP server in listener, conversion-listener, or upstream-server modes. This doesn't affect the passthrough-listener mode as the state in that mode is maintained by the upstream MCP servers.
              - …
            - `timeout` integer: The timeout for calling the tools in milliseconds.
          - `allowed_versions` string[]: The MCP protocol revisions this server accepts. Leave unset to accept every revision Kong implements, which is the default. When set, `server/discover` advertises exactly this list and a request declaring anything else is rejected. Listing only per-request revisions refuses handshake clients. **Requires a minimum runtime version of `2.1`**.
          - `cache` object: Cache hints Kong emits on the cacheable operations it serves. Only clients on a protocol revision that defines them receive them. **Requires a minimum runtime version of `2.1`**.
            - `tools_list` AIGatewayMCPServerCacheHint: A cache hint Kong emits on a cacheable operation it serves.
              - …
            - `discover` AIGatewayMCPServerCacheHint: A cache hint Kong emits on a cacheable operation it serves.
              - …
        - `sources` AIGatewayEntityIdentifier[], required: The explicit list of source MCP Servers whose tools this listener exposes. Each entry is the immutable `name` of a `conversion-only` (toolset) or `upstream-server` (third-party MCP server) MCP Server in the same AI Gateway. All of the referenced source's tools are exposed.
        - `access` union
          - AIGatewayMCPServerListenerConsumer: Auth strategy and OAuth 2.0 Protected Resource Metadata configuration for granting access to an MCP server.
            - `acl_attribute_type` 'consumer', required: The type of attributes that ACL is evaluated with.
            - `acls` union: Access control rules for MCP resources. Configure exactly one of `allow`, `deny`, `allow_when`, or `deny_when`.
              - …
            - `default_tool_acls` AIGatewayMCPACLs: Access control rules for MCP resources. Configure `allow`, `deny`, or both.
              - …
            - `identity_providers` AIGatewayIdentityProviderReference[]: List of identity providers for granting access to the MCP server. At most 1 identity provider of each identity provider type can be referenced. Deprecated: use `auth_strategies` instead. The two are mutually exclusive.
            - `auth_strategies` AIGatewayAuthStrategyReference[]: List of auth strategies for granting access to the MCP server. At most 1 auth strategy of each auth strategy type can be referenced.
            - `metadata` AIGatewayMCPServerProtectedResourceMetadata: OAuth 2.0 Protected Resource Metadata (RFC 9728) advertised for this MCP server, allowing clients to discover the authorization servers that protect it.
              - …
          - AIGatewayMCPServerListenerOauth: Auth strategy and OAuth 2.0 Protected Resource Metadata configuration for granting access to an MCP server.
            - `acl_attribute_type` 'oauth_access_token', required: The type of attributes that ACL is evaluated with.
            - `access_token_claim_field` string, required: The claim in the OAuth2 access token to use as the subject for ACL evaluation when `acl_attribute_type` is set to `oauth_access_token`. Nested claim can be fetched by using a jq filter starts with dot, e.g., “.user.email”: https://jqlang.org/manual/#object-identifier-index
            - `acls` union: Access control rules for MCP resources. Configure exactly one of `allow`, `deny`, `allow_when`, or `deny_when`.
              - …
            - `default_tool_acls` AIGatewayMCPACLs: Access control rules for MCP resources. Configure `allow`, `deny`, or both.
              - …
            - `identity_providers` AIGatewayIdentityProviderReference[]: List of identity providers for granting access to the MCP server. At most 1 identity provider of each identity provider type can be referenced. Deprecated: use `auth_strategies` instead. The two are mutually exclusive.
            - `auth_strategies` AIGatewayAuthStrategyReference[]: List of auth strategies for granting access to the MCP server. At most 1 auth strategy of each auth strategy type can be referenced.
            - `metadata` AIGatewayMCPServerProtectedResourceMetadata: OAuth 2.0 Protected Resource Metadata (RFC 9728) advertised for this MCP server, allowing clients to discover the authorization servers that protect it.
              - …
        - `display_name` string, required: The display name for the MCP Server.
        - `name` string, required: Identifier for an AI Gateway entity. In some cases, this may be the entity name or ID.
        - `enabled` boolean: Whether the MCP Server is enabled.
        - `policies` string[]: List of policy references.
        - `labels` PublicLabels: Public labels store information about an entity that can be used for filtering a list of objects. Public labels are intended to store **PUBLIC** metadata. Keys must be of length 1-63 characters, and cannot start with "kong", "konnect", "mesh", "kic", or "_".
        - `managed_by` ManagedBy: Stores information about what manages this entity, such as the tool or system responsible for its lifecycle (for example, `terraform`). Keys must be 1–63 characters long, start and end with an alphanumeric character, and may contain dashes (`-`), underscores (`_`), and dots (`.`). Values must be 1–127 characters long and must not be empty. Keys and values must not contain spaces or control characters.
        - `datastores` AIGatewayDatastoreRef[]: The Token Vault Datastore holding this entity's credential storage: the directory credentials are written to, and the encryption keys that seal them. Required when `token_vault` is set (on an auth strategy, when its credential manager is enabled), and the referenced Datastore must have type `token-vault`. Only one Datastore can be referenced. **Requires a minimum runtime version of `2.3`**.
          - `name` string, required: An immutable user-defined identifier for this resource.
        - `token_vault` AIGatewayTokenVault: Resolves an upstream credential per request via Kong's Token Vault instead of sending a static credential. The directory credentials live in, and the keys that encrypt them, come from the `token-vault` Datastore this entity references, and are returned read-only in `storage`. Callers must enroll with the configured Token Vault provider before the credential can be resolved. Until enrollment completes, an MCP Server serves virtual `authenticate` and `check_authentication_status` tools that guide the caller through the enrollment flow. **Requires a minimum runtime version of `2.3`**.
          - `provider` string, required: Name of the upstream credential provider registered in the Token Vault directory.
          - `subject_token_type` 'access_token' | 'api_key': The kind of bearer token the caller presents, which the Token Vault exchanges for the upstream credential. Omit it unless the request shape makes the token type ambiguous: it is otherwise resolved from the request, and failing that from the token's shape (a JWS is an `access_token`, anything else an `api_key`).
          - `credential` AIGatewayTokenVaultCredential: Where the caller's subject token is read from, and where the credential the Token Vault returns is placed on the upstream request. Both placements default to `Authorization: Bearer <credential>` in a header, though an `outbound` placement left unset honours whatever the provider advertises instead. Set them for providers that take the credential elsewhere, such as an `x-api-key` header carrying the raw credential.
            - `inbound` AIGatewayTokenVaultCredentialPlacement
              - …
            - `outbound` AIGatewayTokenVaultCredentialPlacement
              - …
          - `cache` object: Optional best-effort cache for exchanged credentials, shared across the cluster. Losing it only costs a re-exchange. Credentials are encrypted with the keys held by the Token Vault Datastore before being cached.
            - `redis` AIGatewayRedisCloudConfiguration: Config for connecting to a Cloud Provider's Redis instance.
              - …
          - `storage` TokenVaultDatastoreConfig
            - `directory` string, required: The name of the directory in which tokens are stored.
            - `current_version` string, required: The version of the encryption key used to encrypt new tokens.
            - `keys` TokenVaultDatastoreKey[], required: The encryption keys used to protect stored tokens. Keep previous versions in the list so existing tokens can still be decrypted.
              - …
      - object
        - `id` string, uuid, required: Contains a unique identifier used for this resource.
        - `created_at` string, date-time, required: An ISO-8601 timestamp representation of entity creation date.
        - `updated_at` string, date-time, required: An ISO-8601 timestamp representation of entity update date.
        - `type` 'passthrough-listener', required
        - `config` AIGatewayMCPServerWithUpstreamConfig, required: Routing, logging, and server configuration for the MCP Server.
          - `route` union, required
            - object: Configuration for an AI Gateway route.
              - …
            - object: Configuration for an AI Gateway route.
              - …
            - object: Configuration for an AI Gateway route.
              - …
            - object: Configuration for an AI Gateway route.
              - …
          - `logging` object: Configuration for AI Gateway logging.
            - `payloads` boolean
            - `audits` boolean
          - `max_request_body_size` integer: Maximum size of request body to parse. Set to 0 for unlimited.
          - `server` object: Server-side configuration for the MCP Server.
            - `forward_client_headers` boolean: Whether to forward the client request headers to the upstream server when calling the tools.
            - `session` object: Enable managed session when Kong responds as MCP server in listener, conversion-listener, or upstream-server modes. This doesn't affect the passthrough-listener mode as the state in that mode is maintained by the upstream MCP servers.
              - …
            - `timeout` integer: The timeout for calling the tools in milliseconds.
          - `url` string, uri, required: Helper field to set protocol, host, port and path of the upstream service using a URL. This is the same as a Kong Gateway Service URL: ${scheme}://${host}:${port}/${path}
          - `upstream` AIGatewayUpstreamConfig: Configuration applied when proxying to the upstream service, including authentication.
            - `auth` AIGatewayUpstreamAuthAWS: AWS IAM (SigV4) authentication for the upstream service.
              - …
          - `certificates` AIGatewayCertificateReference[]: Client certificates used for mutual TLS (mTLS) authentication with the upstream MCP server, referenced by name. Each entry must be the `name` of an existing AI Certificate on the same AI Gateway. The certificates are applied to the Kong Gateway Service generated for this MCP Server, which proxies requests to the upstream MCP server. The generated Service presents a single client certificate (`client_certificate`) during the TLS handshake with the upstream, so at most one certificate may be configured. If unset, no client certificate is presented to the upstream.
          - `ca_certificates` AIGatewayCACertificateReference[]: CA certificates used to verify the TLS certificate of the upstream MCP server, referenced by name. Each entry must be the `name` of an existing AI CA Certificate on the same AI Gateway. The CA certificates build the trust store (`ca_certificates`) of the Kong Gateway Service generated for this MCP Server, which proxies requests to the upstream MCP server. If unset, the default trust store is used to verify the upstream's TLS certificate.
          - `proxy` AIGatewayProxyConfig: HTTP/HTTPS proxy configuration for outbound requests to the upstream AI provider.
            - `http_proxy` object: HTTP proxy server to route plaintext outbound requests through.
              - …
            - `https_proxy` object: HTTPS proxy server to route TLS outbound requests through.
              - …
            - `proxy_scheme` 'http': The proxy scheme to use when connecting to the proxy server.
            - `auth` object: Credentials used to authenticate to the proxy server.
              - …
            - `no_proxy` string: Comma-separated list of hosts that should not be proxied.
        - `tools` AIGatewayMCPPassthroughTool[]: Per-tool access-control overrides for tools advertised by the remote MCP Server. Each entry is matched to a remote tool by `name`; only its access-control rules are applied.
          - `access` AIGatewayMCPToolAccess, required: Access-control rules for a tool.
            - `acls` AIGatewayMCPACLs: Access control rules for MCP resources. Configure `allow`, `deny`, or both.
              - …
          - `name` string, required: Tool identifier used to match remote MCP Server tools for ACL enforcement.
        - `access` union
          - AIGatewayMCPServerListenerConsumer: Auth strategy and OAuth 2.0 Protected Resource Metadata configuration for granting access to an MCP server.
            - `acl_attribute_type` 'consumer', required: The type of attributes that ACL is evaluated with.
            - `acls` union: Access control rules for MCP resources. Configure exactly one of `allow`, `deny`, `allow_when`, or `deny_when`.
              - …
            - `default_tool_acls` AIGatewayMCPACLs: Access control rules for MCP resources. Configure `allow`, `deny`, or both.
              - …
            - `identity_providers` AIGatewayIdentityProviderReference[]: List of identity providers for granting access to the MCP server. At most 1 identity provider of each identity provider type can be referenced. Deprecated: use `auth_strategies` instead. The two are mutually exclusive.
            - `auth_strategies` AIGatewayAuthStrategyReference[]: List of auth strategies for granting access to the MCP server. At most 1 auth strategy of each auth strategy type can be referenced.
            - `metadata` AIGatewayMCPServerProtectedResourceMetadata: OAuth 2.0 Protected Resource Metadata (RFC 9728) advertised for this MCP server, allowing clients to discover the authorization servers that protect it.
              - …
          - AIGatewayMCPServerListenerOauth: Auth strategy and OAuth 2.0 Protected Resource Metadata configuration for granting access to an MCP server.
            - `acl_attribute_type` 'oauth_access_token', required: The type of attributes that ACL is evaluated with.
            - `access_token_claim_field` string, required: The claim in the OAuth2 access token to use as the subject for ACL evaluation when `acl_attribute_type` is set to `oauth_access_token`. Nested claim can be fetched by using a jq filter starts with dot, e.g., “.user.email”: https://jqlang.org/manual/#object-identifier-index
            - `acls` union: Access control rules for MCP resources. Configure exactly one of `allow`, `deny`, `allow_when`, or `deny_when`.
              - …
            - `default_tool_acls` AIGatewayMCPACLs: Access control rules for MCP resources. Configure `allow`, `deny`, or both.
              - …
            - `identity_providers` AIGatewayIdentityProviderReference[]: List of identity providers for granting access to the MCP server. At most 1 identity provider of each identity provider type can be referenced. Deprecated: use `auth_strategies` instead. The two are mutually exclusive.
            - `auth_strategies` AIGatewayAuthStrategyReference[]: List of auth strategies for granting access to the MCP server. At most 1 auth strategy of each auth strategy type can be referenced.
            - `metadata` AIGatewayMCPServerProtectedResourceMetadata: OAuth 2.0 Protected Resource Metadata (RFC 9728) advertised for this MCP server, allowing clients to discover the authorization servers that protect it.
              - …
        - `rate_limiting` AIGatewayMCPServerRateLimiting: Rate limits applied to requests to this MCP Server, matched by MCP method or tool name. Each AI Gateway node counts and enforces the limits independently. **Requires a minimum runtime version of `2.3`**.
          - `limits` AIGatewayMCPServerRateLimit[], required: The rate limits to enforce. Each entry targets exactly one of `method` or `tool`. Two entries cannot share the same target and `window_size`.
            - `method` string: The MCP method this limit applies to.
            - `tool` string: The name of the tool this limit applies to.
            - `limit` integer, required: The number of requests allowed per window.
            - `window_size` 1 | 60 | 3600 | 86400, required: The window size in seconds, one of a second, a minute, an hour, or a day.
        - `display_name` string, required: The display name for the MCP Server.
        - `name` string, required: Identifier for an AI Gateway entity. In some cases, this may be the entity name or ID.
        - `enabled` boolean: Whether the MCP Server is enabled.
        - `policies` string[]: List of policy references.
        - `labels` PublicLabels: Public labels store information about an entity that can be used for filtering a list of objects. Public labels are intended to store **PUBLIC** metadata. Keys must be of length 1-63 characters, and cannot start with "kong", "konnect", "mesh", "kic", or "_".
        - `managed_by` ManagedBy: Stores information about what manages this entity, such as the tool or system responsible for its lifecycle (for example, `terraform`). Keys must be 1–63 characters long, start and end with an alphanumeric character, and may contain dashes (`-`), underscores (`_`), and dots (`.`). Values must be 1–127 characters long and must not be empty. Keys and values must not contain spaces or control characters.
        - `datastores` AIGatewayDatastoreRef[]: The Token Vault Datastore holding this entity's credential storage: the directory credentials are written to, and the encryption keys that seal them. Required when `token_vault` is set (on an auth strategy, when its credential manager is enabled), and the referenced Datastore must have type `token-vault`. Only one Datastore can be referenced. **Requires a minimum runtime version of `2.3`**.
          - `name` string, required: An immutable user-defined identifier for this resource.
        - `token_vault` AIGatewayTokenVault: Resolves an upstream credential per request via Kong's Token Vault instead of sending a static credential. The directory credentials live in, and the keys that encrypt them, come from the `token-vault` Datastore this entity references, and are returned read-only in `storage`. Callers must enroll with the configured Token Vault provider before the credential can be resolved. Until enrollment completes, an MCP Server serves virtual `authenticate` and `check_authentication_status` tools that guide the caller through the enrollment flow. **Requires a minimum runtime version of `2.3`**.
          - `provider` string, required: Name of the upstream credential provider registered in the Token Vault directory.
          - `subject_token_type` 'access_token' | 'api_key': The kind of bearer token the caller presents, which the Token Vault exchanges for the upstream credential. Omit it unless the request shape makes the token type ambiguous: it is otherwise resolved from the request, and failing that from the token's shape (a JWS is an `access_token`, anything else an `api_key`).
          - `credential` AIGatewayTokenVaultCredential: Where the caller's subject token is read from, and where the credential the Token Vault returns is placed on the upstream request. Both placements default to `Authorization: Bearer <credential>` in a header, though an `outbound` placement left unset honours whatever the provider advertises instead. Set them for providers that take the credential elsewhere, such as an `x-api-key` header carrying the raw credential.
            - `inbound` AIGatewayTokenVaultCredentialPlacement
              - …
            - `outbound` AIGatewayTokenVaultCredentialPlacement
              - …
          - `cache` object: Optional best-effort cache for exchanged credentials, shared across the cluster. Losing it only costs a re-exchange. Credentials are encrypted with the keys held by the Token Vault Datastore before being cached.
            - `redis` AIGatewayRedisCloudConfiguration: Config for connecting to a Cloud Provider's Redis instance.
              - …
          - `storage` TokenVaultDatastoreConfig
            - `directory` string, required: The name of the directory in which tokens are stored.
            - `current_version` string, required: The version of the encryption key used to encrypt new tokens.
            - `keys` TokenVaultDatastoreKey[], required: The encryption keys used to protect stored tokens. Keep previous versions in the list so existing tokens can still be decrypted.
              - …
      - object
        - `id` string, uuid, required: Contains a unique identifier used for this resource.
        - `created_at` string, date-time, required: An ISO-8601 timestamp representation of entity creation date.
        - `updated_at` string, date-time, required: An ISO-8601 timestamp representation of entity update date.
        - `type` 'upstream-server', required
        - `config` AIGatewayMCPServerUpstreamServerConfig, required: Routing, logging, and server configuration for the MCP Server.
          - `route` union, required
            - object: Configuration for an AI Gateway route.
              - …
            - object: Configuration for an AI Gateway route.
              - …
            - object: Configuration for an AI Gateway route.
              - …
            - object: Configuration for an AI Gateway route.
              - …
          - `logging` object: Configuration for AI Gateway logging.
            - `payloads` boolean
            - `audits` boolean
          - `max_request_body_size` integer: Maximum size of request body to parse. Set to 0 for unlimited.
          - `server` AIGatewayMCPServerUpstreamServerServerConfig: Server-side configuration specific to `upstream-server` mode.
            - `forward_client_headers` boolean: Whether to forward the client request headers to the upstream server when calling the tools.
            - `session` object: Enable managed session when Kong responds as MCP server in listener, conversion-listener, or upstream-server modes. This doesn't affect the passthrough-listener mode as the state in that mode is maintained by the upstream MCP servers.
              - …
            - `timeout` integer: The timeout for calling the tools in milliseconds.
            - `preserve_upstream_tool_names` boolean: If enabled, the original upstream tool names are preserved as-is when Kong acts as an MCP server. If disabled (`false`), the service name will be prepended to the MCP tool names to avoid name collisions when multiple services are used.
            - `tools_list_auth` union: Configuration for an Upstream Server's MCP Server Tools' Authentication.
              - …
            - `upstream_protocol_version` '2026-07-28' | '2025-11-25' | '2025-06-18' | '2025-03-26': The MCP protocol revision Kong speaks to the upstream MCP server. Leave unset to negotiate a handshake revision with an `initialize` exchange, which is the default. Set a per-request revision to reach an upstream that answers no handshake and mints no session. **Requires a minimum runtime version of `2.1`**.
          - `url` string, uri, required: Helper field to set protocol, host, port and path of the upstream service using a URL. This is the same as a Kong Gateway Service URL: ${scheme}://${host}:${port}/${path}
          - `upstream` AIGatewayUpstreamConfig: Configuration applied when proxying to the upstream service, including authentication.
            - `auth` AIGatewayUpstreamAuthAWS: AWS IAM (SigV4) authentication for the upstream service.
              - …
          - `certificates` AIGatewayCertificateReference[]: Client certificates used for mutual TLS (mTLS) authentication with the upstream MCP server, referenced by name. Each entry must be the `name` of an existing AI Certificate on the same AI Gateway. The certificates are applied to the Kong Gateway Service generated for this MCP Server, which proxies requests to the upstream MCP server. The generated Service presents a single client certificate (`client_certificate`) during the TLS handshake with the upstream, so at most one certificate may be configured. If unset, no client certificate is presented to the upstream.
          - `ca_certificates` AIGatewayCACertificateReference[]: CA certificates used to verify the TLS certificate of the upstream MCP server, referenced by name. Each entry must be the `name` of an existing AI CA Certificate on the same AI Gateway. The CA certificates build the trust store (`ca_certificates`) of the Kong Gateway Service generated for this MCP Server, which proxies requests to the upstream MCP server. If unset, the default trust store is used to verify the upstream's TLS certificate.
          - `tools_cache_ttl_seconds` integer, required: The time-to-live (TTL) for the upstream tools cache in seconds. Set to `0` to refresh on every client call.
        - `tools` AIGatewayMCPUpstreamTool[]: List of tools exposed by this MCP Server.
          - `access` AIGatewayMCPToolAccess: Access-control rules for a tool.
            - `acls` AIGatewayMCPACLs: Access control rules for MCP resources. Configure `allow`, `deny`, or both.
              - …
          - `annotations` AIGatewayMCPToolAnnotations
            - `destructive_hint` boolean: If true, the tool may perform destructive updates
            - `idempotent_hint` boolean: If true, repeated calls with same args have no additional effect
            - `open_world_hint` boolean: If true, tool interacts with external entities
            - `read_only_hint` boolean: If true, the tool does not modify its environment
            - `title` string: Human-readable title for the tool
          - `description` string: A description of what the tool does.
          - `name` string, required: The MCP tool name. In upstream-server mode, it also matches the remote MCP Server tool whose metadata this entry overrides.
          - `input_schema` object, nullable: The entire `inputSchema` section for the tool. Overrides the upstream server's `inputSchema` for the same tool name, if present.
          - `output_schema` object, nullable: The entire `outputSchema` section for the tool. Overrides the upstream server's `outputSchema` for the same tool name, if present.
        - `display_name` string, required: The display name for the MCP Server.
        - `name` string, required: Identifier for an AI Gateway entity. In some cases, this may be the entity name or ID.
        - `enabled` boolean: Whether the MCP Server is enabled.
        - `policies` string[]: List of policy references.
        - `labels` PublicLabels: Public labels store information about an entity that can be used for filtering a list of objects. Public labels are intended to store **PUBLIC** metadata. Keys must be of length 1-63 characters, and cannot start with "kong", "konnect", "mesh", "kic", or "_".
        - `managed_by` ManagedBy: Stores information about what manages this entity, such as the tool or system responsible for its lifecycle (for example, `terraform`). Keys must be 1–63 characters long, start and end with an alphanumeric character, and may contain dashes (`-`), underscores (`_`), and dots (`.`). Values must be 1–127 characters long and must not be empty. Keys and values must not contain spaces or control characters.
        - `datastores` AIGatewayDatastoreRef[]: The Token Vault Datastore holding this entity's credential storage: the directory credentials are written to, and the encryption keys that seal them. Required when `token_vault` is set (on an auth strategy, when its credential manager is enabled), and the referenced Datastore must have type `token-vault`. Only one Datastore can be referenced. **Requires a minimum runtime version of `2.3`**.
          - `name` string, required: An immutable user-defined identifier for this resource.
        - `token_vault` AIGatewayTokenVault: Resolves an upstream credential per request via Kong's Token Vault instead of sending a static credential. The directory credentials live in, and the keys that encrypt them, come from the `token-vault` Datastore this entity references, and are returned read-only in `storage`. Callers must enroll with the configured Token Vault provider before the credential can be resolved. Until enrollment completes, an MCP Server serves virtual `authenticate` and `check_authentication_status` tools that guide the caller through the enrollment flow. **Requires a minimum runtime version of `2.3`**.
          - `provider` string, required: Name of the upstream credential provider registered in the Token Vault directory.
          - `subject_token_type` 'access_token' | 'api_key': The kind of bearer token the caller presents, which the Token Vault exchanges for the upstream credential. Omit it unless the request shape makes the token type ambiguous: it is otherwise resolved from the request, and failing that from the token's shape (a JWS is an `access_token`, anything else an `api_key`).
          - `credential` AIGatewayTokenVaultCredential: Where the caller's subject token is read from, and where the credential the Token Vault returns is placed on the upstream request. Both placements default to `Authorization: Bearer <credential>` in a header, though an `outbound` placement left unset honours whatever the provider advertises instead. Set them for providers that take the credential elsewhere, such as an `x-api-key` header carrying the raw credential.
            - `inbound` AIGatewayTokenVaultCredentialPlacement
              - …
            - `outbound` AIGatewayTokenVaultCredentialPlacement
              - …
          - `cache` object: Optional best-effort cache for exchanged credentials, shared across the cluster. Losing it only costs a re-exchange. Credentials are encrypted with the keys held by the Token Vault Datastore before being cached.
            - `redis` AIGatewayRedisCloudConfiguration: Config for connecting to a Cloud Provider's Redis instance.
              - …
          - `storage` TokenVaultDatastoreConfig
            - `directory` string, required: The name of the directory in which tokens are stored.
            - `current_version` string, required: The version of the encryption key used to encrypt new tokens.
            - `keys` TokenVaultDatastoreKey[], required: The encryption keys used to protect stored tokens. Keep previous versions in the list so existing tokens can still be decrypted.
              - …

## Other responses

- `401`: Unauthorized
- `403`: Forbidden
- `404`: Not Found
- `429`: Too Many Requests

## Changes

- **2026-11-12** `d8bd04cfbaf7`: 131 breaking, 123 warning, 12 info
  - the `acl_attribute_type` response property const value `consumer` was removed for the status `200`
  - the `acl_attribute_type` response property const value `consumer` was removed for the status `200`
  - the `acl_attribute_type` response property const value `consumer` was removed for the status `200`
  - the `acl_attribute_type` response property const value `oauth_access_token` was removed for the status `200`
  - …262 more
- …earlier changes not shown

[Operation history](https://skmtc.dev/kong/apis/konnect-api-go-sdk/changes/v1/ai-gateways/:gatewayId/auth-strategies/:authStrategyIdOrName/usage/get.md)

---

[API](https://skmtc.dev/kong/apis/konnect-api-go-sdk.md) - [All operations](https://skmtc.dev/kong/apis/konnect-api-go-sdk/llms.txt) - [OpenAPI document](https://skmtc.dev/kong/apis/konnect-api-go-sdk/revisions/8184ff111d66?raw)
