@abloatai/ablo 0.24.0 → 0.26.0

package/docs/sessions.md

@@ -53,7 +53,7 @@ return Response.json({ token });  // return ONLY the token to the browser
 A user session has **full data authority** within its org — no operation
 allowlist. It's the human acting as themselves.
 
-Build a browser `Ablo` client whose `getToken` fetches from that route, and pass
+Point a browser `Ablo` client's `authEndpoint` at that route, and pass
 the **instance** to [`<AbloProvider>`](/react). The client fetches the token,
 opens the connection, and re-mints before expiry — your app writes no token
 plumbing:
@@ -67,10 +67,7 @@ import { schema } from '@/ablo.schema';
 
 const ablo = Ablo({
   schema,
-  getToken: () =>
-    fetch('/api/ablo-session', { method: 'POST' })
-      .then((r) => r.json())
-      .then((d) => d.token),
+  authEndpoint: '/api/ablo-session',
 });
 
 export function Providers({ children }: { children: React.ReactNode }) {
@@ -81,8 +78,9 @@ export function Providers({ children }: { children: React.ReactNode }) {
 The client owns auth, the credential lifecycle, and the connection; the provider
 is the thin reactive binding over it (Stripe's `<Elements stripe={...}>` model).
 Build the client **once** at module scope — a new instance per render tears down
-the socket. `authEndpoint: '/api/ablo-session'` is accepted as sugar for the
-`getToken` fetch above if you prefer a URL.
+the socket. Need custom headers or a
+body on the exchange? `authEndpoint` also accepts an async resolver that
+returns the token.
 
 ## Agent sessions (`rk_`)
 

package/llms.txt

@@ -122,9 +122,11 @@ A schema is model fields and relations. Advanced schema helpers such as `mutable
 
 ## Storage Boundary
 
-`databaseUrl` is an OPTIONAL, server-only constructor option on `Ablo(...)`. It is NOT auto-read from the environment — pass it EXPLICITLY to register your Postgres directly. Omit it when you expose a signed Data Source endpoint, or when trying Ablo against the hosted sandbox (apiKey only). A `DATABASE_URL` set for another tool (Prisma, Drizzle, docker-compose) is ignored unless you pass `databaseUrl` explicitly.
+THE ONE PATH IS LOGICAL REPLICATION. In production every schema model is backed by YOUR OWN database, and Ablo connects to it by CONSUMING ITS WRITE-AHEAD LOG (logical replication) — the Electric/PowerSync/Zero model. Ablo tails the WAL and fans changes out; YOUR APPLICATION CONTINUES TO OWN THE WRITE PATH (your own backend, your existing API). Ablo never runs DDL on, writes to, owns, or migrates your database. Connect once with `npx ablo connect` (prints the setup SQL: `wal_level=logical`, a publication, a `REPLICATION` role) then `npx ablo connect --register`. Validate with `npx ablo connect --check`. The role's privilege footprint is narrow but real — a `REPLICATION`-attributed role that streams the WAL and `SELECT`s; do NOT describe it as "read-only" (a security reviewer will push back — it is a replication privilege, just tightly scoped).
 
-In production, every schema model is backed by YOUR OWN database. The PRIMARY path is the connection string: pass `databaseUrl` (most users already have a Postgres — often Prisma- or Drizzle-managed for auth/audit/log tables that are NOT in the Ablo schema; Ablo syncs a SUBSET of models against it). Most users do NOT run `ablo migrate` — they ADOPT existing tables with `npx ablo pull` / `npx ablo check`, or keep managing tables with their own migration tool. Run `npx ablo migrate` only when Ablo should OWN the tables (it provisions the synced-model tables in your DB). The alternative to the connection string is a signed Data Source endpoint that hands Ablo an ORM `adapter` (Drizzle is the default; Prisma and Kysely are also supported) — it owns the transaction, exactly-once idempotency, and outbox in ONE pass (no hand-written `commit`/`events`); use it when database credentials must never leave your infrastructure (or for a local/private-range DB, which Ablo's cloud cannot reach over the network).
+FALLBACK (only when the database can't grant a `REPLICATION` role, e.g. a locked-down managed DB): a signed Data Source endpoint that hands Ablo an ORM `adapter` (Drizzle default; Prisma and Kysely supported) — it owns the transaction, exactly-once idempotency, and outbox in ONE pass; your credentials never leave your infrastructure.
+
+DEPRECATED — being removed, do NOT recommend or scaffold: the `databaseUrl` constructor option (the dial-in, where Ablo held a read/write connection and committed directly to your DB). The end-state constructor is `Ablo({ schema, apiKey })`; the database is connected out of band via `ablo connect`, not through `databaseUrl`. (Note: `ablo init` still defaults to `--storage direct` scaffolding `databaseUrl` — that default is itself slated to lead with `ablo connect`; prefer logical replication regardless of what `init` scaffolds.)
 
 GOTCHA the user WILL hit: `DATABASE_URL` must use a NON-superuser, NON-BYPASSRLS role (Ablo enforces row-level security; owner roles are rejected with `database_role_cannot_enforce_rls`). Neon's and Supabase's default dashboard connection strings use the database OWNER (e.g. `neondb_owner`) and are rejected. EASIEST: `npx ablo migrate` detects the unsafe role and creates the scoped one automatically from the user's machine (owner credential never reaches Ablo; new DATABASE_URL written to the env file). Manual alternative — create a scoped role first: `CREATE ROLE ablo_app LOGIN PASSWORD '...' NOSUPERUSER NOBYPASSRLS; GRANT CREATE, CONNECT ON DATABASE <db> TO ablo_app; GRANT CREATE, USAGE ON SCHEMA public TO ablo_app;` — then swap user/password into the same host/db string.
 

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@abloatai/ablo",
-  "version": "0.24.0",
+  "version": "0.26.0",
   "description": "The Collaboration Layer For AI Agents",
   "license": "Apache-2.0",
   "type": "module",
@@ -108,11 +108,6 @@
       "import": "./dist/server/index.js",
       "default": "./dist/server/index.js"
     },
-    "./server/next": {
-      "types": "./dist/server/next.d.ts",
-      "import": "./dist/server/next.js",
-      "default": "./dist/server/next.js"
-    },
     "./webhooks": {
       "types": "./dist/webhooks/index.d.ts",
       "import": "./dist/webhooks/index.js",
@@ -121,7 +116,6 @@
   },
   "files": [
     "dist",
-    "!dist/internal/**",
     "!dist/__type_probe.*",
     "docs/*.md",
     "docs/examples/*.md",
@@ -139,7 +133,9 @@
     "build": "npm run clean && tsc -p tsconfig.build.json && npm run build:cli",
     "build:cli": "tsup --config tsup.cli.config.ts",
     "typecheck:cli": "tsc -p tsconfig.cli.json",
-    "pack:check": "npm_config_cache=${TMPDIR:-/tmp}/ablo-npm-cache npm pack --dry-run",
+    "prepack": "node scripts/strip-source-condition.mjs",
+    "postpack": "node scripts/restore-source-condition.mjs",
+    "pack:check": "node scripts/pack-check.mjs",
     "lint": "npm run lint:imports && npm run lint:errors && npm run lint:docs && npm run lint:mintlify",
     "build:docs": "node scripts/build-mintlify-docs.mjs",
     "lint:mintlify": "node scripts/build-mintlify-docs.mjs --check",
@@ -147,6 +143,7 @@
     "generate:errors": "tsx scripts/generate-error-docs.mts",
     "lint:errors": "tsx scripts/check-error-docs.mts",
     "lint:docs": "node scripts/check-doc-drift.mjs",
+    "lint:eslint": "eslint . --cache",
     "lint:pkg": "publint",
     "prepublishOnly": "npm run build && npm run lint:pkg",
     "check:dist": "node scripts/check-dist-fresh.mjs",
@@ -191,7 +188,8 @@
   },
   "peerDependencies": {
     "react": "^19.0.0",
-    "drizzle-orm": ">=0.44.0"
+    "drizzle-orm": ">=0.44.0",
+    "ai": "^6.0.0"
   },
   "peerDependenciesMeta": {
     "react": {
@@ -199,6 +197,9 @@
     },
     "drizzle-orm": {
       "optional": true
+    },
+    "ai": {
+      "optional": true
     }
   },
   "dependencies": {
@@ -211,21 +212,21 @@
   "devDependencies": {
     "@ai-sdk/provider": "^3.0.0",
     "@clack/prompts": "^0.11.0",
+    "@jest/globals": "^29.7.0",
     "@prisma/client": "^7.3.0",
     "@types/jest": "^29.5.0",
     "@types/node": "^22.0.0",
     "@types/react": "^19.0.0",
     "@types/uuid": "^10.0.0",
     "@testing-library/react": "^16.0.0",
-    "@testing-library/jest-dom": "^6.6.0",
     "ai": "^6.0.0",
     "drizzle-orm": "^0.45.2",
+    "eslint-plugin-react-hooks": "^7.0.1",
     "fake-indexeddb": "^6.0.0",
     "fast-check": "^3.0.0",
+    "globals": "^16.5.0",
     "jest": "^29.7.0",
     "jest-environment-jsdom": "^29.7.0",
-    "jiti": "^2.7.0",
-    "mobx": "^6.13.7",
     "picocolors": "^1.1.0",
     "postgres": "^3.4.0",
     "react": "^19.0.0",
@@ -233,8 +234,9 @@
     "ts-jest": "^29.4.0",
     "ts-morph": "^26.0.0",
     "tsup": "^8.0.0",
+    "tsx": "^4.19.0",
     "typescript": "^5.8.3",
     "publint": "^0.3.21",
-    "yjs": "^13.6.29"
+    "ws": "^8.18.0"
   }
 }

package/dist/client/index.d.ts

@@ -1,36 +0,0 @@
-/**
- * @abloatai/ablo/client — Consumer API
- *
- * The one-liner entry point for external consumers.
- *
- * `Ablo({ apiKey })` is the stateless HTTP API client. Add `schema`
- * when you want the realtime sync engine with typed model proxies.
- *
- * ```ts
- * import { Ablo } from '@abloatai/ablo/client';
- * import { schema } from './schema';
- *
- * const ablo = Ablo({
- *   schema,
- *   apiKey: process.env.ABLO_API_KEY,
- * });
- *
- * const reports = ablo.weatherReports.list({ where: { status: 'pending' } });
- * await ablo.weatherReports.create({ location: 'Stockholm', status: 'pending' });
- * ```
- *
- * For headless agents (workers, bots), pass the same schema and an API key
- * scoped for that server runtime:
- *
- * ```ts
- * const bot = Ablo({
- *   schema,
- *   apiKey: process.env.ABLO_API_KEY,
- * });
- * ```
- */
-export { Ablo, computeFKDepthPriority, type AbloOptions, type InternalAbloOptions, type ClaimedOptions, type IfClaimedPolicy, type ClaimWaitOptions, type LocalCountOptions, type LocalReadOptions, type ModelListScope, type ServerReadOptions, type ModelOperations, type ModelReadOptions, } from './Ablo.js';
-export { ABLO_DEFAULT_BASE_URL, ABLO_HOSTED_API_DOMAIN, ABLO_HOSTED_HTTP_BASE_URL, normalizeAbloHostedBaseUrl, } from './auth.js';
-export type { AbloPersistence } from './persistence.js';
-export type { AbloApi, AbloApiClientOptions, AbloApiClaims, Capability, CapabilityCreateOptions, CapabilityParticipantKind, CapabilityRecord, CapabilityResource, CapabilityRevocation, CapabilityScope, } from './ApiClient.js';
-export type { EngineParticipant, JoinedParticipant, ParticipantJoinOptions, ParticipantManager, ParticipantScope, ParticipantStatus, ScopedClaims, ScopedPresence, } from '../sync/participants.js';

package/dist/client/index.js

@@ -1,33 +0,0 @@
-/**
- * @abloatai/ablo/client — Consumer API
- *
- * The one-liner entry point for external consumers.
- *
- * `Ablo({ apiKey })` is the stateless HTTP API client. Add `schema`
- * when you want the realtime sync engine with typed model proxies.
- *
- * ```ts
- * import { Ablo } from '@abloatai/ablo/client';
- * import { schema } from './schema';
- *
- * const ablo = Ablo({
- *   schema,
- *   apiKey: process.env.ABLO_API_KEY,
- * });
- *
- * const reports = ablo.weatherReports.list({ where: { status: 'pending' } });
- * await ablo.weatherReports.create({ location: 'Stockholm', status: 'pending' });
- * ```
- *
- * For headless agents (workers, bots), pass the same schema and an API key
- * scoped for that server runtime:
- *
- * ```ts
- * const bot = Ablo({
- *   schema,
- *   apiKey: process.env.ABLO_API_KEY,
- * });
- * ```
- */
-export { Ablo, computeFKDepthPriority, } from './Ablo.js';
-export { ABLO_DEFAULT_BASE_URL, ABLO_HOSTED_API_DOMAIN, ABLO_HOSTED_HTTP_BASE_URL, normalizeAbloHostedBaseUrl, } from './auth.js';

package/dist/config/index.d.ts

@@ -1,10 +0,0 @@
-/**
- * @abloatai/ablo/config — App initialization
- *
- * One-time setup at app boot. Provides DI interface types
- * and the initSyncEngine() function to wire real implementations.
- */
-export { initSyncEngine, resetSyncEngine, isSyncEngineInitialized } from '../context.js';
-export { noopLogger, noopObservability, noopAnalytics, browserOnlineStatus, defaultSessionErrorDetector, emptyConfig, type SyncEngineContext, } from '../SyncEngineContext.js';
-export type { SyncEngineConfig, SyncLogger, SyncObservabilityProvider, SyncAnalytics, MutationExecutor, MutationDispatcher, SessionErrorDetector, OnlineStatusProvider, CommitResult, MutationOperation, BreadcrumbLevel, SyncBreadcrumbCategory, TransactionFailureDetails, BootstrapFailureDetails, WebSocketErrorDetails, RollbackDetails, SpanAttributes, } from '../interfaces/index.js';
-export { SyncSessionError } from '../errors.js';

package/dist/config/index.js

@@ -1,12 +0,0 @@
-/**
- * @abloatai/ablo/config — App initialization
- *
- * One-time setup at app boot. Provides DI interface types
- * and the initSyncEngine() function to wire real implementations.
- */
-// Context lifecycle
-export { initSyncEngine, resetSyncEngine, isSyncEngineInitialized } from '../context.js';
-// Context type + no-op defaults (for testing or gradual adoption)
-export { noopLogger, noopObservability, noopAnalytics, browserOnlineStatus, defaultSessionErrorDetector, emptyConfig, } from '../SyncEngineContext.js';
-// Errors
-export { SyncSessionError } from '../errors.js';

package/dist/interfaces/headless.d.ts

@@ -1,95 +0,0 @@
-/**
- * Headless DI interfaces — the three abstractions createSyncEngine needs
- * to run in Node.js without browser globals.
- *
- * STATUS: Type stubs only. No implementation yet. These interfaces are
- * the deliverable of Path C Phase 1. When all three have both a browser
- * implementation and a Node implementation, the SDK is truly headless.
- *
- * Usage (once implemented):
- *
- *   // Browser (default — no DI needed, uses browser globals)
- *   const engine = createSyncEngine({ url, userId, organizationId });
- *
- *   // Node.js / agent / sidecar (headless — DI overrides)
- *   import { inMemoryStorage, alwaysOnline } from '@abloatai/ablo/headless';
- *   const engine = createSyncEngine({
- *     url, userId, organizationId,
- *     storage: inMemoryStorage(),
- *     network: alwaysOnline(),
- *   });
- *
- * The `transport` override is optional because Node 22 has a built-in
- * global `WebSocket` class. It's included for testing (mock WebSocket)
- * and for environments where the built-in WebSocket isn't available
- * (Node 20, Deno, Bun — each has its own WebSocket story).
- *
- * See also: apps/sync-server/src/sdk-headless-entrypoint.test.ts
- * (the skipped tests that become the acceptance tests for Phase 1)
- */
-/** A record in an object store. */
-export type StorageRecord = Record<string, unknown> & {
-    id: string;
-};
-/** An object store — analogous to an IDBObjectStore. */
-export interface ObjectStore {
-    get(id: string): Promise<StorageRecord | undefined>;
-    getAll(): Promise<StorageRecord[]>;
-    getAllFromIndex(indexName: string, value: string): Promise<StorageRecord[]>;
-    put(record: StorageRecord): Promise<void>;
-    delete(id: string): Promise<void>;
-    clear(): Promise<void>;
-}
-/** A database — analogous to an IDBDatabase. */
-export interface StorageDatabase {
-    getStore(name: string): ObjectStore | undefined;
-    close(): void;
-}
-/**
- * Opens or creates a named database.
- *
- * Browser: wraps indexedDB.open(name, version)
- * Node: returns an in-memory database backed by Maps
- */
-export interface StorageProvider {
-    open(name: string, version?: number): Promise<StorageDatabase>;
-    delete(name: string): Promise<void>;
-}
-export interface NetworkProvider {
-    /** Current online status. */
-    isOnline(): boolean;
-    /**
-     * Subscribe to online/offline transitions. Returns an unsubscribe function.
-     * The callback receives `true` when coming online, `false` when going offline.
-     *
-     * Browser: listens for window 'online'/'offline' events + visibilitychange
-     * Node: never fires (always online)
-     * Test: fires when the test calls setOnline(true/false)
-     */
-    onStatusChange(callback: (online: boolean) => void): () => void;
-}
-/** Minimal WebSocket interface that SyncWebSocket uses. */
-export interface WebSocketLike {
-    readonly readyState: number;
-    readonly bufferedAmount: number;
-    send(data: string | ArrayBuffer): void;
-    close(code?: number, reason?: string): void;
-    ping?(): void;
-    addEventListener(type: 'open', handler: () => void): void;
-    addEventListener(type: 'close', handler: (event: {
-        code: number;
-        reason: string;
-    }) => void): void;
-    addEventListener(type: 'message', handler: (event: {
-        data: unknown;
-    }) => void): void;
-    addEventListener(type: 'error', handler: (event: unknown) => void): void;
-    removeEventListener(type: string, handler: (...args: unknown[]) => void): void;
-}
-export interface TransportProvider {
-    /**
-     * Create a WebSocket connection to the given URL.
-     * Returns an object satisfying the WebSocketLike interface.
-     */
-    connect(url: string): WebSocketLike;
-}

package/dist/interfaces/headless.js

@@ -1,41 +0,0 @@
-/**
- * Headless DI interfaces — the three abstractions createSyncEngine needs
- * to run in Node.js without browser globals.
- *
- * STATUS: Type stubs only. No implementation yet. These interfaces are
- * the deliverable of Path C Phase 1. When all three have both a browser
- * implementation and a Node implementation, the SDK is truly headless.
- *
- * Usage (once implemented):
- *
- *   // Browser (default — no DI needed, uses browser globals)
- *   const engine = createSyncEngine({ url, userId, organizationId });
- *
- *   // Node.js / agent / sidecar (headless — DI overrides)
- *   import { inMemoryStorage, alwaysOnline } from '@abloatai/ablo/headless';
- *   const engine = createSyncEngine({
- *     url, userId, organizationId,
- *     storage: inMemoryStorage(),
- *     network: alwaysOnline(),
- *   });
- *
- * The `transport` override is optional because Node 22 has a built-in
- * global `WebSocket` class. It's included for testing (mock WebSocket)
- * and for environments where the built-in WebSocket isn't available
- * (Node 20, Deno, Bun — each has its own WebSocket story).
- *
- * See also: apps/sync-server/src/sdk-headless-entrypoint.test.ts
- * (the skipped tests that become the acceptance tests for Phase 1)
- */
-export {};
-// ── Factory functions (to be implemented in Phase 1) ─────────────────────
-//
-// These will be the public API that headless consumers import:
-//
-//   import { inMemoryStorage, alwaysOnline } from '@abloatai/ablo/headless';
-//
-// Stubs below show the intended signatures. Implementation is Phase 1 work.
-// export function inMemoryStorage(): StorageProvider { ... }
-// export function alwaysOnline(): NetworkProvider { ... }
-// export function nodeWebSocket(): TransportProvider { ... }
-// export function controllableNetwork(): NetworkProvider & { setOnline(v: boolean): void } { ... }

package/dist/query/index.d.ts

@@ -1,6 +0,0 @@
-/**
- * Query module barrel — re-exports the public surface for convenience.
- * See types.ts and client.ts for the actual definitions.
- */
-export type { Query, QueryBatch, QueryBatchResult } from './types.js';
-export { postQuery, type PostQueryOptions } from './client.js';

package/dist/query/index.js

@@ -1,5 +0,0 @@
-/**
- * Query module barrel — re-exports the public surface for convenience.
- * See types.ts and client.ts for the actual definitions.
- */
-export { postQuery } from './client.js';

package/dist/realtime/index.d.ts

@@ -1,10 +0,0 @@
-/**
- * Internal compatibility entrypoint for the schema-powered realtime client.
- *
- * Use this build for applications that need typed model proxies,
- * subscriptions, presence, offline queueing, and a long-lived WebSocket.
- */
-export { Ablo, computeFKDepthPriority } from '../client/Ablo.js';
-export type { AbloOptions, InternalAbloOptions, LocalCountOptions, LocalReadOptions, ModelListScope, ServerReadOptions, ModelOperations, } from '../client/Ablo.js';
-import { Ablo } from '../client/Ablo.js';
-export default Ablo;

package/dist/realtime/index.js

@@ -1,9 +0,0 @@
-/**
- * Internal compatibility entrypoint for the schema-powered realtime client.
- *
- * Use this build for applications that need typed model proxies,
- * subscriptions, presence, offline queueing, and a long-lived WebSocket.
- */
-export { Ablo, computeFKDepthPriority } from '../client/Ablo.js';
-import { Ablo } from '../client/Ablo.js';
-export default Ablo;

package/dist/schema/plane.d.ts

@@ -1,23 +0,0 @@
-/**
- * Database PLANE — which database a model's rows live in. A sibling axis to
- * `tenancy` (which says how rows are isolated *within* a database):
- *
- *   - `tenant`  — the tenant data plane. For a BYO/dedicated customer this is
- *                 THEIR database; provisioning emits these tables there.
- *   - `control` — Ablo's control plane (the sync log, attribution, audit, …).
- *                 Never emitted into a customer DB; lives only in Ablo's own DB.
- *
- * P1 of the sync-delta decomposition (`docs/plans/sync-delta-zod-decomposition.md`):
- * declaring the boundary lets BYO provisioning *derive* "what a customer DB gets"
- * (`plane === 'tenant'`) instead of hand-coding it. Defaults to `tenant` —
- * today every `defineSchema` model is the customer's own data; only Ablo's
- * internal tables (once modeled in P2) declare `control`.
- */
-import { z } from 'zod';
-export declare const planeSchema: z.ZodEnum<{
-    tenant: "tenant";
-    control: "control";
-}>;
-export type SchemaPlane = z.infer<typeof planeSchema>;
-/** Default plane for a model that doesn't declare one — the tenant data plane. */
-export declare const DEFAULT_PLANE: SchemaPlane;

package/dist/schema/plane.js

@@ -1,19 +0,0 @@
-/**
- * Database PLANE — which database a model's rows live in. A sibling axis to
- * `tenancy` (which says how rows are isolated *within* a database):
- *
- *   - `tenant`  — the tenant data plane. For a BYO/dedicated customer this is
- *                 THEIR database; provisioning emits these tables there.
- *   - `control` — Ablo's control plane (the sync log, attribution, audit, …).
- *                 Never emitted into a customer DB; lives only in Ablo's own DB.
- *
- * P1 of the sync-delta decomposition (`docs/plans/sync-delta-zod-decomposition.md`):
- * declaring the boundary lets BYO provisioning *derive* "what a customer DB gets"
- * (`plane === 'tenant'`) instead of hand-coding it. Defaults to `tenant` —
- * today every `defineSchema` model is the customer's own data; only Ablo's
- * internal tables (once modeled in P2) declare `control`.
- */
-import { z } from 'zod';
-export const planeSchema = z.enum(['tenant', 'control']);
-/** Default plane for a model that doesn't declare one — the tenant data plane. */
-export const DEFAULT_PLANE = 'tenant';

package/dist/server/next.d.ts

@@ -1,51 +0,0 @@
-/**
- * `@abloatai/ablo/server/next` — mount Ablo's HTTP surface in a consumer's
- * Next.js App Router, the Better-Auth `toNextJsHandler` way:
- *
- * ```ts
- * // app/api/ablo/[...all]/route.ts
- * import { toAbloHandler } from '@abloatai/ablo/server/next';
- * import { ablo } from '@/lib/ablo';        // your configured Ablo HTTP app
- * export const { GET, POST } = toAbloHandler(ablo);
- * ```
- *
- * Ablo's HTTP endpoints (bootstrap / query / the commit fallback) are stateless
- * request→response, so they drop straight into a route handler. The realtime
- * WebSocket channel is deliberately NOT served here — it is a persistent
- * connection a serverless route handler cannot host, and stays an Ablo-hosted
- * (or long-running Node) service. That is the HTTP/WSS split the architecture
- * draws on purpose (see docs/plans/sync-gateway-audit-log-architecture.md).
- *
- * The package owns this MOUNT PRIMITIVE; the host owns the infra-bound app it
- * wraps (the Hono app holding the Hub + routes + auth). Mirrors Better Auth,
- * where `betterAuth()` builds the `.handler` and `toNextJsHandler` just maps it
- * onto the HTTP verbs.
- */
-/**
- * Anything that resolves a Web `Request` to a `Response` — e.g. a Hono app's
- * `.fetch`, a Better-Auth-style `{ handler }`, or a bare function.
- */
-export type AbloFetchHandler = (request: Request) => Response | Promise<Response>;
-/** The shapes `toAbloHandler` accepts: a Hono-style `{ fetch }`, a
- *  `{ handler }`, or a bare `(Request) => Response`. */
-export type AbloHttpApp = {
-    readonly fetch: AbloFetchHandler;
-} | {
-    readonly handler: AbloFetchHandler;
-} | AbloFetchHandler;
-/** The route-handler object a Next.js App Router `route.ts` re-exports. One
- *  handler bound to every method the `[...all]` catch-all may receive. */
-export interface AbloRouteHandlers {
-    readonly GET: AbloFetchHandler;
-    readonly POST: AbloFetchHandler;
-    readonly PATCH: AbloFetchHandler;
-    readonly PUT: AbloFetchHandler;
-    readonly DELETE: AbloFetchHandler;
-    readonly OPTIONS: AbloFetchHandler;
-}
-/**
- * Wrap a configured Ablo HTTP app into Next.js App Router route handlers.
- * Mirrors Better Auth's `toNextJsHandler` — accepts the app's `fetch`/`handler`
- * (or a bare function) and returns one handler per verb.
- */
-export declare function toAbloHandler(app: AbloHttpApp): AbloRouteHandlers;

package/dist/server/next.js

@@ -1,47 +0,0 @@
-/**
- * `@abloatai/ablo/server/next` — mount Ablo's HTTP surface in a consumer's
- * Next.js App Router, the Better-Auth `toNextJsHandler` way:
- *
- * ```ts
- * // app/api/ablo/[...all]/route.ts
- * import { toAbloHandler } from '@abloatai/ablo/server/next';
- * import { ablo } from '@/lib/ablo';        // your configured Ablo HTTP app
- * export const { GET, POST } = toAbloHandler(ablo);
- * ```
- *
- * Ablo's HTTP endpoints (bootstrap / query / the commit fallback) are stateless
- * request→response, so they drop straight into a route handler. The realtime
- * WebSocket channel is deliberately NOT served here — it is a persistent
- * connection a serverless route handler cannot host, and stays an Ablo-hosted
- * (or long-running Node) service. That is the HTTP/WSS split the architecture
- * draws on purpose (see docs/plans/sync-gateway-audit-log-architecture.md).
- *
- * The package owns this MOUNT PRIMITIVE; the host owns the infra-bound app it
- * wraps (the Hono app holding the Hub + routes + auth). Mirrors Better Auth,
- * where `betterAuth()` builds the `.handler` and `toNextJsHandler` just maps it
- * onto the HTTP verbs.
- */
-function resolveFetch(app) {
-    if (typeof app === 'function')
-        return app;
-    if ('fetch' in app)
-        return app.fetch;
-    return app.handler;
-}
-/**
- * Wrap a configured Ablo HTTP app into Next.js App Router route handlers.
- * Mirrors Better Auth's `toNextJsHandler` — accepts the app's `fetch`/`handler`
- * (or a bare function) and returns one handler per verb.
- */
-export function toAbloHandler(app) {
-    const fetch = resolveFetch(app);
-    const handler = (request) => fetch(request);
-    return {
-        GET: handler,
-        POST: handler,
-        PATCH: handler,
-        PUT: handler,
-        DELETE: handler,
-        OPTIONS: handler,
-    };
-}

package/dist/sync/OfflineFlush.d.ts

@@ -1,9 +0,0 @@
-/**
- * OfflineFlush — Replays queued offline mutations on reconnect.
- *
- * SDK-generic version: delegates to MutationDispatcher from context.
- */
-export declare function flushOfflineQueueOnce(): Promise<{
-    processed: number;
-    failed: number;
-}>;

package/dist/sync/OfflineFlush.js

@@ -1,22 +0,0 @@
-/**
- * OfflineFlush — Replays queued offline mutations on reconnect.
- *
- * SDK-generic version: delegates to MutationDispatcher from context.
- */
-import { OfflineTransactionStore } from './OfflineTransactionStore.js';
-import { getContext } from '../context.js';
-let _offlineTxStore = null;
-function getOfflineTxStore() {
-    if (!_offlineTxStore) {
-        _offlineTxStore = new OfflineTransactionStore();
-    }
-    return _offlineTxStore;
-}
-export async function flushOfflineQueueOnce() {
-    const store = getOfflineTxStore();
-    await store.init();
-    const dispatcher = getContext().mutationDispatcher;
-    return store.flush(async (tx) => {
-        await dispatcher.dispatch(tx.opName, tx.request.variables || {});
-    });
-}

package/dist/sync/OfflineTransactionStore.d.ts

@@ -1,37 +0,0 @@
-export declare enum Priority {
-    CRITICAL = 0,
-    HIGH = 1,
-    NORMAL = 2,
-    LOW = 3
-}
-export type OfflineTransaction = {
-    id: string;
-    opName: string;
-    priority: Priority;
-    createdAt: number;
-    dependsOn?: string[];
-    request: {
-        gqlId?: string;
-        document?: string;
-        variables?: Record<string, any>;
-        headers?: Record<string, string>;
-        method?: string;
-        url?: string;
-        body?: any;
-    };
-};
-export declare class OfflineTransactionStore {
-    private db;
-    private key;
-    init(): Promise<void>;
-    enqueue(tx: Omit<OfflineTransaction, 'createdAt'>): Promise<string>;
-    listAll(): Promise<OfflineTransaction[]>;
-    getOptimizedSyncOrder(): Promise<OfflineTransaction[]>;
-    remove(id: string): Promise<void>;
-    clear(): Promise<void>;
-    flush(processor: (tx: OfflineTransaction) => Promise<void>): Promise<{
-        processed: number;
-        failed: number;
-    }>;
-}
-export declare const offlineTxStore: OfflineTransactionStore;