Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.auvy.ai/llms.txt

Use this file to discover all available pages before exploring further.

​
Assets and Artifacts API

Assets and artifacts are the public document lanes for AUVY memory:
LaneEndpointBody policyUse for
Assets/v1/assetsImmutable body; metadata/path can changeUploaded files, ingested text, URL imports, RAG-backed knowledge
Artifacts/v1/artifactsMutable UTF-8 body with content versionsGenerated docs, notes, summaries, editable Markdown/HTML/plain text
Both lanes are stored on the shared resource substrate, appear in Vault browse, and are discoverable through Search (hybrid keyword + vector) and Recollections.
Send X-Brain-Id or configure SDK getBrainId for brain-scoped reads and writes. /v1/assets rejects artifact rows and /v1/artifacts rejects asset rows.

​
SDK Namespace

The public SDK uses resourceStore for these lanes: 
const assetRows = await auvy.resourceStore.list({ lane: 'assets', limit: 20 })
const artifactRows = await auvy.resourceStore.list({ lane: 'artifacts', limit: 20 })

const { resource, embeddingJobId } = await auvy.resourceStore.ingest({
  source: { type: 'text', value: 'Support refund rules...' },
  name: 'Refund policy',
  path: '/Policies',
})

const artifact = await auvy.resourceStore.createArtifactRow({
  title: 'Customer briefing',
  content: '# Briefing\n\nLatest account context...',
  content_type: 'text/markdown',
})

​
Assets

GET    /v1/assets
POST   /v1/assets
POST   /v1/assets/ingest
GET    /v1/assets/:id
PUT    /v1/assets/:id
DELETE /v1/assets/:id
Create an asset directly with metadata, or ingest text, a URL, or multipart file data. 
curl -X POST https://api.auvy.ai/v1/assets/ingest \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "X-Brain-Id: YOUR_BRAIN_ID" \
  -H "Content-Type: application/json" \
  -d '{
    "source": { "type": "text", "value": "Refund policies and escalation notes..." },
    "name": "Refund policy",
    "path": "/Policies"
  }'
Updating an asset can change fields such as name, metadata, path, or skip_embedding. It cannot replace the stored body; ingest a new asset instead.

​
Artifacts

GET    /v1/artifacts
POST   /v1/artifacts
POST   /v1/artifacts/patch
GET    /v1/artifacts/:id
PUT    /v1/artifacts/:id
DELETE /v1/artifacts/:id
GET    /v1/artifacts/:id/body
Artifacts are text-first and versioned. Use them when agents or users need to revise a document. 
curl -X POST https://api.auvy.ai/v1/artifacts \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "X-Brain-Id: YOUR_BRAIN_ID" \
  -H "Content-Type: application/json" \
  -d '{
    "title": "Customer briefing",
    "content": "# Briefing\n\nLatest account context...",
    "content_type": "text/markdown"
  }'
Fetch raw artifact body with optional version selection: 
GET /v1/artifacts/:id/body?version=2&format=markdown

​
Chunks and Versions

Both lanes expose chunk and version helpers: 
GET    /v1/assets/:id/chunks
GET    /v1/artifacts/:id/chunks
GET    /v1/assets/chunks/:chunk_id
GET    /v1/artifacts/chunks/:chunk_id
GET    /v1/assets/:id/versions
GET    /v1/artifacts/:id/versions
DELETE /v1/assets/:id/chunks
DELETE /v1/assets/:id/chunks/by-source
Assets generally record metadata versions. Artifacts record content versions.

​
Folder Structure

GET  /v1/assets/structure
GET  /v1/artifacts/structure
POST /v1/assets/structure-batch
POST /v1/resources/tree
POST /v1/resources/move
Use path, path_prefix, and the resource tree APIs to keep Vault folder views in sync.
  • Resources — unified resource catalog, Vault tree, and live events
  • Browse — list-only resource discovery
  • Search — hybrid search over indexed document-lane chunks
  • Recollections — owner-scoped retrieval over assets and artifacts