Fork Database

Scoped persistent storage

Every active space can use Fork’s managed document database. The service derives its scope from trusted network identity, so application code never submits a workspace ID or database credential. The default namespace is per space; deployments may intentionally configure project- or organization-level sharing.

fork-data scope
fork-data tables
fork-data --help

Use the CLI

Use the preinstalled CLI for inspection, seed data, migrations, and one-off changes.

fork-data create-table contacts
fork-data insert contacts '{"name":"Ada","status":"lead"}'
fork-data list contacts --limit 100
fork-data get contacts DOCUMENT_ID
fork-data patch contacts DOCUMENT_ID '{"status":"active"}'

Documents are JSON objects. Lists return an opaque next_cursor; pass it back with --afterfor the next page.

Use it from your backend

Server code reads the base URL from FORK_DATA_API_URL. Keep calls behind application-owned routes that validate the signed-in user and return only the records that user may access.

const base = process.env.FORK_DATA_API_URL;
if (!base) throw new Error("Fork Database is unavailable");

const response = await fetch(`${base}/tables/contacts/documents`, {
  method: "POST",
  headers: { "Content-Type": "application/json" },
  body: JSON.stringify({ name: "Ada", status: "lead" }),
});

if (!response.ok) throw new Error(`Fork Database ${response.status}`);
const document = await response.json();

Inspect data in the control panel

People with control access can open the space’s read-only Database view to list tables and page through JSON documents. The browser never receives the underlying service URL or a database credential.