Component

core/security/nemo-guardrails

NVIDIA NeMo Guardrails' server as a guard in a model's path: a service speaking the OpenAI chat-completions API that checks each request against its input rail, forwards what it admits to the upstream model API in infra.upstream_endpoint, and returns the upstream's answer, streamed or not. A request the rail refuses is answered by the guard itself with an ordinary completion, and nothing of it reaches the upstream. The rail sends the latest user message to the detector wired on the detector socket and refuses the request when the detector reports one of values.blocked_labels at or above values.threshold; a detector that cannot be reached, or answers with an error, refuses the request too. It judges the latest user message alone. Its verdict is a probabilistic classifier's: it misses some attacks, refuses some legitimate requests and is known to be evadable. Only a flow in its configuration, redefining the product's own, keeps a request body from switching the rail off, and that closure has to be proved again for any image or flow it was not proved on. NVIDIA's NemoGuard models add rails on the content_safety and jailbreak_detection sockets; the jailbreak rail admits a request when its classifier is unreachable. It authenticates no caller, and with no credential attached to its upstream_key slot a caller can reach the upstream on a credential of its own. The Release supplies the image in infra.image and the catalog names none: TrustyAI's published build carries NLTK's punkt tokenizer model, whose licence is unresolved upstream, and is published only under a floating tag, whose superseded digests stop resolving. Nothing renders until the Release sets `infra.image`.

Provenance

OfficialSigned by a signing identity this catalog recognizes as an official publisher.
Content digest
sha256:f871e845d92afd93e2d6d8c7fc6a9b33f86abbcfe1730cc89aab9d44fb3f2721
Published
2026-11-10T10:13:39Z
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

    sha256:4a2b09d9b13c5c2efc29928cc9bd5abc2304b72043e199a4b9ffddab6eb8b06e

Reference at this version

oci://hub.aggrega.ai/component/core/security/nemo-guardrails:1.0.0

Developer parameters

Values the developer sets in the Aggregate.

NameTypeDefaultRequiredDescription
thresholdnumber (min 0, max 1)0.5noProbability at or above which a label the detector reports refuses the request. Measured with the detector's default model, which scores most texts close to 0 or close to 1: at 0.5 the guard refuses plain attempts to override a model's instructions; at 0.001 it also refuses an override respelled in leetspeak, which the model scores near 0.008, and with it some ordinary requests the model scores low but not near zero. At either value it refuses the ordinary requests the model scores close to 1, and serves attacks it scores close to 0. Those two numbers hold for that model only, and are measured again for any other.
blocked_labelsarray (min 1)INJECTIONnoLabels of the detector's classifier that refuse a request when reported at or above the threshold. The default is the label the detector's default model gives an attempt to override a model's instructions; another model labels by its own vocabulary, which this list follows. It may not be empty, because a guard refusing on no label forwards everything.

Operator sizing

Infrastructure sizing the operator tunes per environment.

NameTypeDefaultRequiredDescription
portinteger (min 1024, max 65535)8000noPort the guard listens on and its Service accepts callers on. Most operators should omit this and let the Recipe assign the default (8000), the port the published image listens on unless told otherwise.
upstream_endpointstring—yesBase URL of the OpenAI-compatible model API the guard forwards the requests it admits to, with its API path, for example http://model-server:8000/v1. A deterministic pointer string the guard resolves when it forwards; the compiler never contacts it. The request is forwarded under the model name the caller asked for.
imagestring—yesImage the guard runs, supplied by the Release; the catalog names none. The Recipes run the published build's own start-up and paths, so the image is TrustyAI's NeMo Guardrails server or a copy of it. That build carries a data package whose licence is unresolved upstream and is tagged only as latest, under which the registry stops serving a build once the tag moves on, so a Release that must keep one build mirrors it and names the copy here. The guard's refusal of a request-body switch was proved on one build, and is proved again for any other image.
computeobject—no—

Output contract

Values this component publishes to consumers over the uses topology.

NameTypeDefaultRequiredDescription
hoststring—yesIn-cluster DNS name of the Service fronting the guard. Every caller reaching this name can send requests through the guard, so it is an address for the component routing model traffic through it rather than one to publish.
portinteger—yesTCP port that Service accepts callers on. It follows the port an operator set in the Release, so a consumer reads it rather than assuming a default.
endpointstring—yesBase URL of the OpenAI-compatible API the guard serves, with the host, the port and the API path already composed. A consumer hands it to a client library as its base URL, and sends chat completions to it.
protocolstring—noURL scheme to reach `host` and `port` over, for a consumer composing its own address rather than reading `endpoint`.
thresholdnumber—yesvalues.threshold, the probability at or above which a label the detector reports refuses the request.

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.

detector

Selector
security.aggrega.ai/text-detection=detector
Inject
—

This socket declares no attachment slots, so a connection through it carries no attachment.

content_safety

Selector
component.aggrega.ai/category=ai-model, model.aggrega.ai/api=openai
Inject
—

This socket declares no attachment slots, so a connection through it carries no attachment.

jailbreak_detection

Selector
component.aggrega.ai/category=security
Inject
—

This socket declares no attachment slots, so a connection through it carries no attachment.

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.

SlotSelectorRequiredInjectSensitiveDescription
upstream_keyfeature.aggrega.ai/kind=secretnoenvyesOptional credential the guard presents to the upstream model API as a bearer token. Attached, every forwarded request carries it, whatever the caller sends. Unattached, the guard holds no upstream credential, which suits an upstream that asks for none: the product then sends a fixed placeholder in its place, which an upstream requiring a key refuses, or passes a caller's X-Authorization header on as the credential when the caller sends one.

Self-inject allowlist

No self-inject pipes declared.

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

Used by

Component concept reference →

The Aggrega catalog — an open index of components, features, recipes, and the plugins that render them.

Built with the shared Aggrega design system.

Privacy