> ## Documentation Index
> Fetch the complete documentation index at: https://docs.vegalake.com/llms.txt
> Use this file to discover all available pages before exploring further.

# VegaGraph overview

> Understand VegaLake's typed, tenant-scoped metadata graph and its product boundaries.

# VegaGraph

VegaGraph stores typed facts and evidence about data, software, infrastructure, business, governance, and operational resources. It combines a reusable type catalog with workspace-scoped entities, versioned aspects, relationships, contexts, labels, search documents, and lineage.

## What the graph contains

* **Entity types** define the identity and display shape for a class of resource.
* **Aspect types** define reusable metadata schemas attached to entities.
* **Relationship types** constrain valid edge endpoints and metadata.
* **Entities** are tenant-owned resource records.
* **Aspects** add versioned facts to an entity.
* **Edges** connect entities or field references through a relationship type.
* **Contexts** represent environment or perspective overlays.
* **Labels** classify resources for discovery and governance.

Every tenant-owned graph row is scoped to an organization and workspace. Permissions apply to graph collections and individual objects.

## Product areas

| Area         | Use                                                               |
| ------------ | ----------------------------------------------------------------- |
| Overview     | See a workspace-level summary of graph resources                  |
| Explore      | Search and filter entities, then open details                     |
| Entity 360   | Inspect identity, aspects, versions, hierarchy, and relationships |
| Lineage      | Traverse upstream/downstream entity and field dependencies        |
| Contexts     | View facts through default or hierarchical environment scopes     |
| Type Catalog | Understand the schemas and constraints used by the graph          |

## Boundaries

VegaGraph stores durable graph facts; ingestion and agent execution are outside VegaGraph. Producers must call its APIs or another integration layer to write facts. Secret values are references only in graph metadata—never store a credential inside an aspect.

Start with the [VegaGraph quickstart](/vegagraph/quickstart), then learn the [graph model](/vegagraph/model).