maravilla-workflows
Maravilla Workflows
Workflows are durable, replay-based functions. They run inside the runtime, persist every step's output to a ledger, and survive process restarts: on resume, the runtime replays the workflow function up to the last completed step, then continues from there.
This makes them ideal for:
- Multi-step processes where each step depends on the previous (
step.runfor at-most-once side effects) - Long sleeps (
step.sleep('1h'),step.sleepUntil(date)) — the function isn't running while it sleeps - Waiting for an external event (
step.waitForEvent) - Composition (
step.invoketo call a child workflow)
Workflows are unconditionally enabled — there's no opt-in flag. They have no HTTP trigger; you start them from your runtime code via platform.workflows.start(...).
Defining a workflow
Workflow files live in workflows/*.ts (auto-discovered by the framework adapter at build time). Use defineWorkflow from the runtime subpath:
import { defineWorkflow } from '@maravilla-labs/functions/workflows/runtime';
More from maravilla-labs/maravilla-cli
maravilla-auth
Maravilla Cloud authentication. Use whenever wiring login/register/session, OAuth callbacks, resource policies, or hitting `platform.auth.*` APIs. Critical: the 3-step request-scoped contract (validate → setCurrentUser → can) — skipping any step silently breaks Layer-2 policies and owner-scoped reads return empty with no error.
16maravilla-events
Maravilla Cloud event handlers — files in `events/*.ts` auto-discovered by the framework adapter. Use to react to data changes (`onKvChange`, `onDb`), auth lifecycle (`onAuth`), schedule (`onSchedule`), queue messages (`onQueue`), realtime publishes (`onChannel`), deploy phases (`onDeploy`), object storage (`onStorage`), or arbitrary REN events (`defineEvent`). Run inside the Maravilla runtime with full platform access via `ctx`.
16maravilla-media-transforms
Async media + document derivations via `platform.media.transforms` and the declarative `transforms` block in `maravilla.config.ts`. Media: transcode video, thumbnail extraction, image resize/variants, OCR. Documents (.docx/.odt/.pptx/.xlsx/...): convert to PDF, render page thumbnails, generic format conversion, Markdown extraction (RAG-ready), single-file HTML with inlined images, image-replacement templating ({{TAG}} swap + named-object swap), QR-code injection. Use when ingesting user uploads that need normalised renditions, generating contracts/invoices from templates, or extracting structured content for LLMs. Critical: derived keys are content-addressed — `keyFor(srcKey, spec)` is known up front, before the worker starts, so clients can render placeholder UI without round-trips. Declarative config is the default; imperative `transforms.*` calls are for one-offs.
16maravilla-db
Maravilla Cloud document database — MongoDB-style queries, secondary indexes, and vector search. Use for structured app data, multi-field queries, sorting, semantic search via `findSimilar` / hybrid `find` with `options.vector`. Exposed as `platform.env.DB`. Vector indexes support int8/bit quantization, matryoshka, and multi-vector (ColBERT) out of the box.
15maravilla-config
The `maravilla.config.ts` declarative project file. Use whenever creating or modifying auth resources, groups, relations, registration fields, OAuth providers, password/session policy, branding, database indexes, or media transforms. Reconciled into delivery on every deploy — partial adoption is supported (omit a section to leave it untouched).
15maravilla-push
Maravilla Cloud Web Push — server `platform.push.send/schedule/cancelScheduled/listScheduled` with idempotent keys and recurring `everySeconds`, browser `registerPush({ topics, userId, swPath })` from `@maravilla-labs/platform/push`. Use for browser notifications, scheduled reminders, recurring digests, and per-user fan-out by topic.
15