Dataset Passport — Developer Quickstart

Dataset Passport packages Sovereign IDs, ledger anchoring, and Guardian distribution so institutions can prove custody and policy for every dataset.

Use this if you are…

Museums, archives, labs, or ministries publishing datasets to partners and AI labs.
Any steward who needs to prove provenance, storage integrity, and Accord obligations for sensitive data.

What you get

Canonical manifests for datasets, storage pointers, refresh cadence, and custodians.
Accord-bound passports that express AI training, cultural, and commercial rules.
Guardian node publication so labs can subscribe to verified feeds.
Entitlements: 500 dataset slots, Guardian publication enabled, runtime integrity scoring.

Core APIs & SDK touchpoints

Capability	REST Endpoint	SDK hint
Create steward identity	`POST /api/v1/identity`	`client.createIdentity()`
Register dataset	`POST /api/v1/ai/datasets/register`	`client.request('POST', '/ai/datasets/register', …)`
Issue dataset passport	`POST /api/v1/passport`	`client.createPassport()`
Publish Guardian alert/report	`POST /api/v1/guardian/reports`	`client.request('POST', '/guardian/reports', …)`
Verify dataset & integrity score	`POST /api/v1/provenance/verify`	`client.provenance.verifyArtifact()`

SDK quickstart

import { createSovProvClient } from '@sovprovenance/sdk-js'

const client = createSovProvClient({
  baseUrl: process.env.SOVEREIGN_API_URL!,
  apiKey: process.env.SOVEREIGN_API_KEY!, // Map this key to dataset_passport via PRODUCT_SUBSCRIPTIONS
})

// 1. Register the dataset steward identity
const steward = await client.createIdentity({
  name: 'Global Research Library',
  type: 'institution',
  publicKey: 'did:key:z6MkDatasetKey#1',
  jurisdiction: 'EU-DE',
})

// 2. Register the dataset manifest (AI routes accept canonical dataset payloads)
const dataset = await client.request('POST', '/ai/datasets/register', {
  stewardId: steward.id,
  datasetName: 'Carbon Flux Archive',
  storage: {
    type: 's3',
    uri: 's3://carbon-flux/raw/',
    checksum: 'sha256:abc123',
  },
  metadata: {
    refreshCadence: 'monthly',
    schemaVersion: '1.2.0',
  },
})

// 3. Issue a passport bound to Accord rules
const passport = await client.createPassport({
  identityId: steward.id,
  artifactId: dataset.dataset_id,
  metadata: {
    title: 'Carbon Flux Archive',
    artifactType: 'dataset',
    tags: ['climate', 'restricted-ai'],
  },
  accordId: 'accord-carbon-flux',
})

// 4. Publish a Guardian report so subscribers can ingest
await client.request('POST', '/guardian/reports', {
  type: 'dataset_passport',
  audience: ['labs', 'cities'],
  passportId: passport.id,
  summary: 'New climate dataset available under restricted AI terms.',
})

Implementation notes

Storage bindings can reference S3, GCS, IPFS CIDs, or archival IDs—mirror checksums for each shard so Guardian nodes can validate.
Use Accord metadata to express AI training allowances, derivatives, and jurisdictional limits; these become machine-readable policies for labs.
Guardian publication requires enableGuardianNode entitlements (already on for Dataset Passport). Use /guardian/reports to push alerts or /guardian/status to check mirror health.