Personal Data Graph
Personal Data Graph is the relationship-first view of a user's OS objects. It
Position
Personal Data Graph is the relationship-first view of a user's OS objects. It
belongs under /app-user because it is user-owned context rather than an
OS-level capability map.
It answers:
How are my people, projects, prompts, decisions, docs, payments, deployments,
connectors, domains, and Spaces related?Data Graph Versus Data Room
- Data Graph: relationship-first, provenance-first, graph exploration.
- Data Room: asset, file, dataset, and room-first management surface.
This milestone does not modify /app-user/data-room.
Current Surface
- Route:
/app-user/data-graph - Package:
@prox-os/user-apps - Source:
packages/apps/user-apps/src/apps/data-graph/ - Runtime kind:
os-package - Current status: mock UI with typed entities and relationships
The mock model includes people, projects, apps, prompts, AI questions,
decisions, documents, domains, deployments, subscriptions, payments,
connectors, and Spaces. Relationship verbs include asked, created,
belongs_to, deployed_to, documented_by, paid_for, synced_from,
referenced_by, decided_in, and connected_via.
Future Contract
Future production work should connect the graph through:
- app contract data scopes;
- permission-aware connector sync;
- Activity and audit events;
- prompt/action registry metadata;
- workspace and Space membership;
- backend graph or read-model APIs.
Sensitive nodes must be redacted or summarized before AI context use. External connector data should only enter the graph after explicit authorization and auditable sync.
Mock Boundary
No graph backend, database table, connector sync, payment integration, or provider API exists in this milestone.