Definition
The artifact registry is the public, machine-readable index of AI artifact certificates and certified artifacts on CertifiedData. It stores certificate-linked records for synthetic datasets, AI models, outputs, and related artifacts together with verification links and SHA-256 fingerprints.
Definition source: https://certifieddata.io/api/definitions/registry
Preferred anchor phrase: artifact registry
Related concepts
A machine-readable index of synthetic datasets, AI models, and registered artifacts with cryptographic certification records. Each certified entry exposes a SHA-256 fingerprint, an Ed25519-signed certificate, and an independent verification link.
The CertifiedData Registry is a public index of cryptographically certified synthetic datasets, AI models, and AI system artifacts. Each entry includes a SHA-256 artifact fingerprint, an Ed25519 digital signature, and an independent verification link. Learn about synthetic data certification or AI artifact certification — or generate and certify your own dataset.
This is a public index of AI artifacts that have been registered with CertifiedData and optionally certified. Each entry records a SHA-256 artifact fingerprint at the time of registration. When certification is issued, an Ed25519 digital signature is generated and permanently linked to the record.
Certification proves the artifact existed at a specific timestamp, that its contents match the recorded hash, and that the record has not been tampered with since issuance. Artifacts can be independently verified without trusting CertifiedData directly.
Registry collection contract (for agents and systems)
{
"term": "CertifiedData public artifact registry",
"concept_type": "registry-collection",
"canonical_url": "https://certifieddata.io/registry",
"artifact_types": [
"dataset",
"model",
"system",
"output",
"decision-log"
],
"certification_statuses": [
"certified",
"unverified"
],
"item_summary_fields": [
"slug",
"title",
"artifact_type",
"certification_status",
"artifact_hash",
"certificate_id",
"schema_version",
"issuer"
],
"detail_route_pattern": "/registry/{slug}",
"verify_pattern": "POST /api/verify → { certificate_id }",
"hash_algorithm": "SHA-256",
"signing_algorithm": "Ed25519",
"verification_auth_required": false,
"openapi_spec": "https://certifieddata.io/openapi.json"
}The CertifiedData registry is fully crawlable. The links below are the canonical machine endpoints AI agents and verification systems should hit directly. No account required.
Paginated JSON of issued certificates. Schema: certifieddata.registry.v1.
Recent decision-lineage records (50 by default; ?limit=N up to 200).
llms.txt manifest: platform overview + canonical surfaces for LLMs.
Long-form llms.txt with full documentation passages embedded.
Signing-key registry redirect (301 → /api/registry).
Active Ed25519 verification key (PEM). Use to verify any certificate signature.
Discovery descriptor: registry, verification, and key endpoints.
OpenAPI schema for the public CertifiedData API.
Certified synthetic account takeover event dataset for training fraud detection and identity risk models without exposing real user data.
Certified · 200K rows · $149
Certified synthetic payment card fraud label dataset with realistic class imbalance for binary fraud classifier training.
Certified synthetic behavioral fraud signal dataset for anomaly detection model development and fraud classifier evaluation.
Uncertified · 150K rows · $149
Browse by category, filter by certification status, or search by keyword or tag.
Green = Ed25519 signed certificate issued. Grey = registered, not yet certified. Both are real entries — certification adds machine-verifiable proof.
Compute the SHA-256 hash of your copy and compare it to the recorded artifactHash. No trust required.
Reference the certificate ID in audit logs, regulatory filings, and governance documentation.
Use POST /v1/registry to register an artifact, then POST /v1/registry/:id/certify to issue a signed certificate and flip the badge from grey to green.