Cette page a été traduite automatiquement. Lire l’original en anglais.

Component

core/datastore/cognee

A semantic-memory store for agent deployments. It ingests text and documents and turns them into a typed graph of entities and the relations between them, kept alongside embeddings of the same material, so a consumer asks what relates to what rather than only what looks similar. Construction is the point: a model reads each ingested item, names the entities and relations it finds, and the result is written into the graph, optionally grounded against an ontology the deployment supplies so extracted names resolve onto terms the deployment already uses. The graph, the embeddings and the bookkeeping tables are held by this component itself, on its own persistent volume, which is why it is a stateful server rather than a façade over someone else's store. The model it reads with is wired in over the uses topology and its address is a deterministic pointer string the running server resolves per request; the compiler never contacts a model. Realized on Kubernetes via its Helm Recipe, which runs the official upstream image directly and carries the whole memory configuration — the graph engine, the embedding store, the extraction and embedding models, and the ontology grounding — into the rendered artifact. Exports its host, its port, its base URL and its protocol to consumers via the uses topology.

Provenance

OfficielSigné par une identité de signature que ce catalogue reconnaît comme éditeur officiel.
Empreinte du contenu
sha256:4bdd8f4f621c576eefbce67fb2a33f88a5cf55cb32f9d46beeeae96c5ba66294
Publié
2026-11-10T10:10:28Z
Identité de signature
https://token.actions.githubusercontent.com/https://github.com/aggrega-ai/aggrega-hub/.github/workflows/catalog-publish-sign.yml@refs/heads/main

Signature

  • Officielapplication/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

    sha256:978ba6e470686d5cdd8098768482297abf1b488a9a0b5142d5872c40f11b286f

Référence à cette version

oci://hub.aggrega.ai/component/core/datastore/cognee:1.0.0

Paramètres du développeur

Valeurs que le développeur définit dans l’Aggregate.

NomTypeValeur par défautObligatoireDescription
extraction_providerstringopenainonThe provider family the server reads ingested material with. Consulted when the extraction_model socket is unwired; a wired socket supplies the address and model name instead.
embedding_providerstringopenainonThe provider family the server encodes ingested material with. Consulted when the embedding_model socket is unwired; a wired socket supplies the address and model name instead.
embedding_dimensionsinteger (min 1)—nonWidth of the vectors the embedding model returns. Set it to match the model in use; omit it and the server takes the width its own default model returns.
ontology_file_pathstring—nonPath inside the container to an ontology document the deployment supplies, against which extracted names are resolved onto terms the deployment already uses. A deterministic pointer string the running server opens; the compiler never reads the file. Omit it and extraction proceeds ungrounded, naming entities as it finds them.
extraction_rate_limit_per_minuteinteger (min 1)—nonLargest number of model requests the server issues per minute while building the graph, which is what keeps a large ingestion inside a provider's own quota. Omit it and the server issues requests as fast as ingestion produces them.
access_control_enabledbooleanfalsenonKeep each dataset's memory readable only by the users granted it, rather than serving one shared memory to every caller. Off by default, matching the server's own default.

Dimensionnement de l’opérateur

Dimensionnement de l’infrastructure que l’opérateur ajuste pour chaque environnement.

NomTypeValeur par défautObligatoireDescription
portinteger (min 1024, max 65535)8000nonHTTP listen port the API serves on. Most operators should omit this and let the Recipe assign the default (8000).
engine_versionstring1.5.3nonServer container image tag.
computeobject—non—
storageobject—non—
overridesobject (open)—nonProvider-specific overrides. Accepts any nested structure. Each Recipe documents which keys it reads.

Contrat de sortie

Valeurs que ce composant publie vers les consommateurs via la topologie uses.

NomTypeValeur par défautObligatoireDescription
hoststring—ouiIn-cluster DNS name of the Service fronting the memory API. A consumer that composes its own address dials this name; one that does not reads `endpoint` instead.
portinteger—ouiTCP port that Service accepts API requests on. It follows the port an operator set in the Release, so a consumer reads it rather than assuming a default.
endpointstring—nonFull base URL of the memory API, with the host and the port already composed. A consumer hands it to its client as the address it ingests material through and asks its questions of.
protocolstring—nonHow the published address is spoken to — `http`, so a consumer reaches the API with an HTTP client.

Sockets uses

Les sockets de dépendance par lesquels ce composant se connecte. La cible de chaque connexion câblée via un socket doit satisfaire son sélecteur d’étiquettes, et chaque connexion peut porter une attache dans les emplacements que le socket déclare. Un emplacement obligatoire doit être rempli par chaque connexion que l’Aggregate ou la Release câble via le socket.

extraction_model

Sélecteur
component.aggrega.ai/category in (ai-model, gateway), model.aggrega.ai/api
Injection
—

Emplacements que chaque connexion passant par ce socket peut remplir

EmplacementSélecteurObligatoireInjectionSensibleDescription
credentialfeature.aggrega.ai/kind=secretnonenvouiOptional credential the server presents to the model this connection reaches, the one it reads ingested material with. When attached, the Recipe delivers it as LLM_API_KEY (a Kubernetes secretKeyRef on helm), so its value never appears in the rendered manifest, and the component's model_api_key is not read for this side; unattached, no key is delivered for it.

embedding_model

Sélecteur
component.aggrega.ai/category in (ai-model, gateway), model.aggrega.ai/api
Injection
—

Emplacements que chaque connexion passant par ce socket peut remplir

EmplacementSélecteurObligatoireInjectionSensibleDescription
credentialfeature.aggrega.ai/kind=secretnonenvouiOptional credential the server presents to the model this connection reaches, the one it encodes ingested material with. When attached, the Recipe delivers it as EMBEDDING_API_KEY (a Kubernetes secretKeyRef on helm), so its value never appears in the rendered manifest, and the component's model_api_key is not read for this side; unattached, no key is delivered for it, and the server may then present the extraction side's key, where one is delivered, to this model instead.

Emplacements d’attache

Points de liaison nommés auxquels une Feature s’attache à ce composant. Le sélecteur d’un emplacement restreint la Feature qui peut le remplir, et sa liste inject indique comment le matériel de la Feature peut être livré. Un emplacement obligatoire doit être lié sur chaque instance, par une attache sur l’instance ou à une portée supérieure, et un emplacement sensible est déclaré pour traiter du matériel secret. Une Feature peut aussi s’attacher sous un nom qu’aucun emplacement ne déclare.

EmplacementSélecteurObligatoireInjectionSensibleDescription
model_api_keyfeature.aggrega.ai/kind=secretnonenvouiOptional credential the server presents to a model provider it reaches without a connection. It serves only a model socket left unwired, where the server reaches the provider its values name at that provider's own address: for extraction it is delivered as LLM_API_KEY, for embedding as EMBEDDING_API_KEY, so with both unwired each side presents this key. A wired socket carries its key on its own connection, under credential, and never reads this slot. Delivered as an environment variable, so the plaintext never appears in the rendered manifest; unattached, no key is delivered for an unwired side.

Liste d’autorisation d’auto-injection

env

Relations

Arêtes typées vers et depuis d’autres entités publiées. Une référence sans cible publiée s’affiche comme une arête rompue plutôt que comme un lien.

Utilise

Utilisé par

Référence du concept Component →

Le catalogue Aggrega — un index ouvert de composants, de fonctionnalités, de recettes et des plugins qui en assurent le rendu.

Conçu avec le système de design partagé d’Aggrega.

Confidentialité