Component
core/gateway/agentgateway
agentgateway, an open-source gateway for model, tool and agent traffic, serving the OpenAI chat-completions API in front of the models wired to it. It terminates client HTTP traffic on its data-plane port, publishes that port's host, port and protocol, and runs from one configuration file. Each model wired to its models socket is served under the connection's alias: a request names the alias as its model at the gateway's one /v1 base URL, and is sent to the endpoint and model name the model publishes, presenting that connection's credential. It publishes that base URL as its endpoint and values.model as its model, so an agent's model socket takes it in place of a model; with nothing wired it serves no model and both are empty. A Release may name its callers in infra.callers, their keys through the caller_keys slot; every model request must then present a declared key as `Authorization: Bearer <key>`. Two realizations run the same image, pinned by digest. On Kubernetes, `agentgateway-chart` renders the project's standalone Helm chart, fetched from the project's registry, and realizes data masking, prompt deny patterns, the period budget, audit logging and the retry policy; the fallback chain, target selection, the semantic cache, response validation and the prompt-injection shield halt the build there, each naming why. Under Docker Compose, `agentgateway` runs it as one service and realizes none of them, nor callers, so an attachment halts the build naming its type. With `values.metrics_enabled` true each Recipe publishes the address of the gateway's metrics, always on, on a port that checks no caller, as `metrics_endpoint`; false, that key is empty.
Provenance
- Content digest
- sha256:f5492abed6926aaa7a2eb24792e5d8ce4e330c6bb38abce3fcf062b0ddab7d1d
- Published
- 2026-11-10T10:11:36Z
- Signing identity
- https://token.actions.githubusercontent.com/https://github.com/aggrega-ai/aggrega-hub/.github/workflows/catalog-publish-sign.yml@refs/heads/main
Signature
- Officialapplication/vnd.dev.sigstore.bundle.v0.3+json
https://token.actions.githubusercontent.com/https://github.com/aggrega-ai/aggrega-hub/.github/workflows/catalog-publish-sign.yml@refs/heads/main
sha256:aa49f49c4376940a1cd50e5e440b56616e25622891d38e12189dbe8eb33246db
Reference at this version
oci://hub.aggrega.ai/component/core/gateway/agentgateway:1.0.0Developer parameters
Values the developer sets in the Aggregate.
| Name | Type | Default | Required | Description |
|---|---|---|---|---|
| model | string | — | no | The alias of a connection on the models socket that this gateway publishes as its model, the model name a consumer wired to it sends. It must be an alias a models connection carries, and the build halts naming this value otherwise; it is required while the models socket holds a connection, and omitted while it holds none. |
| metrics_enabled | boolean | false | no | Publishes the address of the gateway's own metrics, in the Prometheus format, as `metrics_endpoint`, putting their port on a Service where a realization needs one; while false that key is empty. The metrics port checks no caller, so anything that reaches it can read them. |
Operator sizing
Infrastructure sizing the operator tunes per environment.
| Name | Type | Default | Required | Description |
|---|---|---|---|---|
| http_port | integer (min 1024, max 65535) | 4000 | no | Data-plane port consumers connect to. |
| callers | array (min 1) | — | no | The callers the gateway serves models to, by name, each presenting its own key in the Authorization header as `Bearer <key>`, which is how an OpenAI client presents one. Every model request then has to present a declared key — one presenting none, or an unknown one, is refused with HTTP 401 — and the key is removed from the request the gateway forwards. A name takes letters, digits, hyphens and underscores, and is also the data key holding that caller's key in the Secret the caller_keys slot names. Read by `agentgateway-chart`; the compose realization halts the build on it. Omit it for open routes. |
| storage | object | — | no | — |
| source_version | string | — | no | Version of the standalone Helm chart the chart-backed realization pulls. The realization holds the chart versions its configuration is proven at, and the build halts naming them on any other. The compose realization pulls no chart and ignores it. |
| compute | object | — | no | — |
| overrides | object (open) | — | no | Provider-specific overrides. Accepts any nested structure. Each Recipe documents which keys it reads. |
Output contract
Values this component publishes to consumers over the uses topology.
| Name | Type | Default | Required | Description |
|---|---|---|---|---|
| host | string | — | yes | DNS name of the Service, or the compose service, fronting the gateway's data plane, the name alone without a namespace. A consumer sends its client traffic here rather than to the model behind the gateway. |
| port | integer | — | yes | TCP port that address accepts model traffic on, infra.http_port. |
| protocol | string | — | no | How the published address is spoken to — `http`, so a consumer composes an address on that scheme. |
| endpoint | string | — | no | Full base URL of the OpenAI-compatible API the gateway serves, with the host, the port and /v1 composed. A caller hands it to a client library as its base URL and names values.model as its model; every other wired model is served at the same base URL under its own alias. Empty while nothing is wired to the models socket. |
| model | string | — | no | The alias a consumer wired to the gateway sends as its model name: values.model. The gateway sends the request on asking for the model name the wired model publishes. Empty while nothing is wired to the models socket. |
| metrics_endpoint | string | — | no | The full address Prometheus scrapes this component's metrics at — scheme, host, port and path, `http://<host>:15020/metrics`. Empty while values.metrics_enabled is false. |
Uses sockets
The dependency sockets this component connects through. The target of every connection wired through a socket must satisfy its label selector, and each connection may carry an attachment in the slots the socket declares. A required slot is owed by each connection the Aggregate or the Release wires through the socket.
models
- Selector
- component.aggrega.ai/category=ai-model
- Inject
- —
Slots each connection through this socket may fill
| Slot | Selector | Required | Inject | Sensitive | Description |
|---|---|---|---|---|---|
| credential | feature.aggrega.ai/kind=secret | no | env | yes | Optional credential the gateway presents to the model this connection reaches, as a bearer token. When attached, the Recipe delivers it as AI_PROXY_<ALIAS>_API_KEY, the connection's alias in capitals with every character other than a letter or a digit as an underscore — a Kubernetes secretKeyRef to the Secret the credential publishes on the chart realization, which halts the build when it publishes none, and the value the credential publishes under that name on the inject pipe on compose — and the connection's model entry names $AI_PROXY_<ALIAS>_API_KEY, which agentgateway replaces from its environment, never the value. The key must be a plain token — letters, digits and - . _ ~ + / = — because agentgateway reads the file as YAML after putting the key into it, and other characters can load as a different key or stop the gateway. Unattached, the entry adds no credential. |
Attachment slots
Named binding points at which a Feature attaches to this component. A slot's selector constrains which Feature may fill it and its inject list names how the Feature's material may be delivered. A required slot must be bound on every instance, by an attachment on the instance or at a scope above it, and a sensitive slot is declared to handle secret material. A Feature may also attach under a name no slot declares.
| Slot | Selector | Required | Inject | Sensitive | Description |
|---|---|---|---|---|---|
| caller_keys | feature.aggrega.ai/kind=secret | no | env | yes | The Secret holding the key each caller infra.callers declares presents to the gateway, one data key per caller, the data key being the caller's name. The attached credential must publish the Secret's name, as core/referenced-secret-set does, and a credential publishing none halts the build. The chart Recipe delivers each key as CALLER_<NAME>_KEY, the caller's name in capitals with every character other than a letter or a digit as an underscore, by a Kubernetes secretKeyRef, and the configuration names $CALLER_<NAME>_KEY, never the key. Each key must be a plain token — letters, digits and - . _ ~ + / = — because agentgateway reads the file as YAML after putting the key into it, and other characters can load as a different key or stop the gateway. Required while infra.callers declares a caller, and refused while it declares none. A caller with no data key of its name in the Secret leaves the gateway unable to start, since Kubernetes starts no container whose secretKeyRef names a missing key; that is caught at start, not by the build. A request presenting an empty key is refused. A key changed inside the Secret takes effect when the gateway's pod next starts, so the key it replaces is accepted until then. |
Self-inject allowlist
env
Relationships
Typed edges to and from other published entities. A reference with no published target renders as a broken edge rather than a link.
Uses
- usesuses socket «models»
component.aggrega.ai/category=ai-model
Attachment slots: credential
core/ai-model/ai21 v1.0.0 →core/ai-model/anthropic v1.0.0 →core/ai-model/azure-managed v1.0.0 →core/ai-model/azure-openai v1.0.0 →core/ai-model/azure-openai-embedding v1.0.0 →core/ai-model/bedrock v1.0.0 →core/ai-model/bedrock-embedding v1.0.0 →core/ai-model/cohere v1.0.0 →core/ai-model/databricks v1.0.0 →core/ai-model/docker-model-runner v1.0.0 →core/ai-model/gemini v1.0.0 →core/ai-model/huggingface v1.0.0 →core/ai-model/litellm v1.0.0 →core/ai-model/mistral v1.0.0 →core/ai-model/nemoguard-content-safety v1.0.0 →core/ai-model/nvidia-api v1.0.0 →core/ai-model/nvidia-nim v1.0.0 →core/ai-model/nvidia-nim-operator v1.0.0 →core/ai-model/ollama v1.0.0 →core/ai-model/openai v1.0.0 →core/ai-model/openai-compatible v1.0.0 →core/ai-model/openai-embedding v1.0.0 →core/ai-model/replicate v1.0.0 →core/ai-model/sagemaker v1.0.0 →core/ai-model/tei v1.0.0 →core/ai-model/tgi v1.0.0 →core/ai-model/vertex-ai v1.0.0 →core/ai-model/vertex-ai-embedding v1.0.0 →core/ai-model/vllm v1.0.0 →
Used by
uses socket «model»
uses socket «model»
uses socket «model»
uses socket «model»
uses socket «model»
uses socket «model»
uses socket «model»
uses socket «model»
uses socket «model»
uses socket «model»
uses socket «model»
uses socket «model»
uses socket «model»
uses socket «model»
uses socket «model»
uses socket «model»
uses socket «model»
uses socket «model»
uses socket «model»
uses socket «model»
uses socket «model»
uses socket «model»
uses socket «model»
uses socket «model»
uses socket «model»
uses socket «model»
uses socket «model»
uses socket «model»
uses socket «model»
uses socket «model»
uses socket «model»
uses socket «model»
uses socket «model»
uses socket «target_model»
uses socket «extraction_model»
uses socket «embedding_model»
uses socket «upstream»
uses socket «forward_target»
uses socket «scrape_target»
uses socket «store»
applies to
applies to
applies to
applies to
applies to
applies to
applies to
applies to
applies to
applies to
applies to
applies to
applies to
applies to
applies to
applies to
applies to
implements
implements