Getting Started

The primary integration path is:

1. generate a backend adapter manifest from e2ee-client-backend
2. point e2ee-backend-adapter at that manifest
3. configure DATABASE_URL
4. run the Rust adapter server

Example:

e2ee-backend-adapter-server \
  --manifest ./generated/e2ee-backend.manifest.json \
  --database-url postgres://postgres:postgres@localhost:5432/app

The current runtime validates the manifest, connects to PostgreSQL, creates the auth tables it needs if they are missing, and exposes adapter metadata endpoints. When the exported expected schema uses --api graphql, the runtime exposes the configured GraphQL endpoint with database-backed entity CRUD plus the GraphQL auth operations. When the exported expected schema uses --api rest, the generated REST entity routes are also backed by the same database CRUD layer.

Before wiring this into a larger project, read File Lifecycle. That page explains which files you author, which files are generated, who consumes each one, and which ones should not be edited by hand.

For generated client schema files, use the CLI export command:

e2ee-backend-adapter-cli export-expected-schema \
  --db-schema-config ./e2ee-backend.db-schema.json \
  --encrypted-schema-config ./e2ee-backend.encrypted-schema.json \
  --api graphql \
  --out ./generated/expected-schema.json \
  --typescript-out ./generated/e2ee-client-bindings.ts

The DB schema config is the structural source of truth for export generation. It is intended to be generated from the database and contains entity names, table mappings, inferred field shapes, and encrypted field placeholders. The encrypted schema config is the user-authored overlay for encrypted field structure and API naming overrides. Use --api graphql or --api rest to choose the client binding surface to export.

If you want a DB schema config from a live Postgres database, generate it with the CLI:

e2ee-backend-adapter-cli generate-db-schema-config \
  --database-url postgres://postgres:postgres@localhost:5432/app \
  --name my-backend \
  --out ./e2ee-backend.db-schema.json

The DB scaffold can infer tables, columns, primary keys, and encrypted ciphertext/nonce pairs, but it cannot know the final logical client shape you want for decrypted objects. That richer structure belongs in e2ee-backend.encrypted-schema.json.

For database reconciliation workflows, use diff, which now writes SQL by default:

e2ee-backend-adapter-cli diff \
  --manifest ./generated/e2ee-backend.manifest.json \
  --database-url postgres://postgres:postgres@localhost:5432/app \
  --out ./generated/schema-diff.sql

If you prefer a SeaORM migration scaffold instead of plain SQL:

e2ee-backend-adapter-cli diff \
  --format seaorm \
  --manifest ./generated/e2ee-backend.manifest.json \
  --database-url postgres://postgres:postgres@localhost:5432/app \
  --out ./migration/src/m20260503_000001_sync_manifest.rs