Whenever

Build and publish Whenever workflow automations from your coding agent. Read the SDK and integration catalogue, validate workflow code, and create workflows in your account. OAuth sign-in.

Hosted MCP Server

npx add-mcp 'https://mcp.whenever.dev'

Installs into Claude Code, Codex, Cursor and more

Documentation

AI writes the automation. It runs the same way, every time.

From a sentence to a running workflow

Describe it in plain words. It asks only what it needs to know.

  • Any workflow, in plain wordsIf you can explain it, it can be built. No node limits, no step caps.
  • Free weekly AI buildingdescribe it, change it, test it; weekly AI usage limits apply
  • 1,400+ integrations, plus any APInot listed? point Whenever at its API docs and it works the calls out
  • Costs cents to run, not AI creditsAI builds the workflow once. After that, runs do not call an AI model, and Pro includes 10,000 runs a month. See the pricing page for current US dollar prices.
  • It’s yoursexport it any time, and any developer can pick it up and run it elsewhere
  • Your canvas or oursbuild it in chat here, or from the coding agent you already use

Coming from: Cursor or Claude

1,402 integrations in the catalogue, plus any other API you can name

Not in the list? Name the service and what you want it to do, and Whenever builds the integration for it.

Browse all integrations

Pay for AI once, not on every run.

AI builds your workflow for free, within weekly usage limits. After that, every run is plain code, with no AI model to pay for.

An AI agent

Calls an AI model on every run

~$880 USD/month* in AI model costs for 10,000 runs

Whenever

Uses AI once, to build it

Loading price… for 10,000 runs, billed yearly

*Estimate: 10,000 runs a month at the per-run token volumes we see on one of Wix's own AI agents, priced at Claude Opus 5.5's published standard rate of $4 per million input tokens and $20 per million output (Anthropic, checked 28 September 2026). Real agents carry more context than this and cost more. An AI step you add to your own workflow draws on your plan's AI allowance on the runs where it happens. All prices are in USD.

Same input. Same result. Every run.

Would you let an AI agent approve a refund unwatched? Here, AI writes the workflow, you approve it once, and every run follows it exactly.

4 different answers

Same request. A new judgement call every run.

Whenever workflow

The same answer, 4 times

Approve it once. Count on it every run.

Start from a brief that already works

Real automations, not demos. Pick the one closest to what you need and Whenever writes it against your accounts; the code is yours to change from there.

Bring every lead source into one CRM recordCollects new leads from Typeform and a Google Sheets tab and creates or updates a single HubSpot contact with its original source.Build this workflow Announce closed-won deals in SlackCelebrates each won deal in Slack with the numbers and a short recap of how it was won.Build this workflow Publish Instagram posts from a calendarPublishes due rows from your Instagram calendar on the cadence you choose and writes each post link back.Build this workflow Log survey responses and alert on low scoresAdds each survey response to Google Sheets within minutes and emails the owner when the score falls below your threshold.Build this workflow Track keyword rankings in a sheet each weekChecks the search position of every keyword on your list, appends the history to a sheet and reports the moves.Build this workflow Route large expense claims to a managerLogs every expense submission in a sheet and alerts a manager in Slack when the amount is over the approval limit.Build this workflow

Browse all templates

Name the tool nobody else supports

Whenever reads the API documentation, tells you which pages it took the calls from, and writes them. Nothing published? Paste the documentation you do have into the chat and it works from that instead.

Whenever a new order lands, push it into

Reading its API documentation…

Documentation found

Source

Its developer reference, 4 pages read

Auth

API key in a header

Paging

Page numbers, 50 per page

Add a secret value

Create that credential in its own settings and paste it here in Setup. It never goes in the chat.

CLINIC_CRM_API_KEY

Save

From there it is ordinary code. Every run makes the same call the same way. And when there is no documentation to find and none to give it, Whenever tells you and changes nothing. It never guesses an endpoint.

Build it in Whenever, or from the coding agent you already use

Chat here, drive it from Claude or Cursor over our MCP server, or write the module yourself. Same runtime, same workflow.

Build with Whenever

Describe the automation in the chat. Whenever writes the workflow code, runs it against real inputs, and shows it to you before anything goes live. No boilerplate, no SDK to learn.

When a Stripe payment over $500 lands, post it in #revenue with the customer name.

Written, with Stripe and Slack connected. Want to read the code or test it on a real payment?

Sign up

Use your coding agent

Already have a coding agent? Paste the prompt below and point it at our MCP server, and it writes, tests and schedules the workflow code on your behalf.

Paste this into your coding agent

When a Stripe payment succeeds, post the
customer, the amount and a link to #revenue.
Use https://www.whenever.dev/skill.md

Then add our MCP to your own coding agent. MCP setup

https://mcp.whenever.dev

Build with code

A workflow is a typed TypeScript module. Write it yourself, keep it in your repo, and let Whenever handle the triggers and the credentials.

import { defineWorkflow, defineStep, event, type WorkflowContext, type WorkflowManifest } from '@wix/whenever-workflow-sdk';

type Payment = { amount: number };

export const manifest: WorkflowManifest = {
  name: 'Post new payments to Slack',
  triggers: [event({ key: 'payment', source: 'stripe.payment-intent-succeeded' })],
};

export const post = defineStep(
  'Post to Slack',
  async (ctx: WorkflowContext<Payment>, amount: number) => {
    await ctx.integrations.slack.postMessage({
      channelId: ctx.config.SLACK_CHANNEL_ID,
      text: \`Payment: ${amount}\`,
    });
  },
);

export default defineWorkflow<Payment, { posted: boolean }>(async (ctx) => {
  if (ctx.trigger?.source !== 'stripe.payment-intent-succeeded') {
    return { posted: false };
  }
  await post(ctx, ctx.input.amount);
  return { posted: true };
});

Every export, trigger and step: read the docs

Join the beta

Easy to follow. Free to leave.

Every workflow shows up as a clear, step-by-step flow. Behind it is a real code file you can download any time, or hand to any developer to run elsewhere.

What Whenever hands you

  • Shopify TriggerWhen an ==order is paid==

  • Prepare message Calculate ==tax== and prepare the message

  • SlackPost the update to ==#ops-orders==

  • See it. Every step, in plain words.

  • Change it. Describe the change, and AI updates it.

  • Export it. Download the real workflow file any time.

  • Keep it. Any developer can run it elsewhere, no Whenever account needed.

What a workflow builder hands you

The same automation, wired by hand, with every field mapped.

workflow.json

Exports only run inside the same builder.

Free to start. Pay per run.

Start with one free workflow, no credit card. Upgrade when you need more workflows or runs. No per-step or per-item counting: one run is one run, however many steps it has.

Hobby

$0/mo

Get access to:

  • 100 runs** monthly
  • One active workflow, no credit card.

Start for free

Paid plans from

Loading price…

Upgrade as you go for more workflows, features and support:

  • 10,000 runs** monthly
  • In-app code edits

Find the right plan for you

* One run = one workflow execution, start to finish. Free AI building and editing have weekly usage limits.

Or bring one over and stop tracing it node by node

Describe the automation you already have and Whenever rebuilds it as one file you read top to bottom instead of a canvas you follow with your finger, testable before it goes live.

Coming from n8nDescribe what your node graph does and Whenever rebuilds it as a module you read top to bottom.Coming from ZapierDescribe your Zap’s triggers and actions and get a workflow you can edit, test and keep.Copy from any* AI agentChatGPT task, Claude, Lindy, Relevance AI, Zapier — anything that runs on a prompt. Paste the prompt in and Whenever writes the workflow from it; you connect the services it needs and test it before it goes live.*Ask your agent to write out the prompt that defines it, or copy its main prompt yourself.

The things people ask first

Do I have to write code?

No. You describe what you want in plain language and Whenever writes the workflow. The code is there when you want it, to read, to review, or to edit, and ignoring it is a perfectly normal way to use Whenever.

What does "workflow-as-code" mean?

A language model writes your automation once, as a typed TypeScript module. Every run after that just executes that code, so run one thousand behaves exactly like run one.

What happens when an API changes?

Nothing changes shape under you: the workflow keeps running the code it was given. When a call starts failing you get told which step broke, and you can ask Whenever to rewrite that step against the new API.

How do triggers work?

A workflow runs on a webhook, on a cron schedule, or on demand. Credentials are managed for you and there is nothing to host, so there is no server to babysit between runs.

What happens if a run fails?

The run is kept along with its inputs and its output, so you can see which step broke and what it was given rather than guessing from a log line. Whenever does not silently re-run it for you: retries and duplicate handling belong in the workflow, and you can ask Whenever to write them.

Can I export my workflows?

Yes. A workflow is a TypeScript module, not a diagram locked inside a database, so you can take it with you.

What does it cost to run?

Nothing calls a model just to run your workflow. The model work happens while the workflow is being written, and AI building is free within weekly usage limits shared across your workflows. After that a run is ordinary code execution, metered separately. If the workflow you wrote calls an AI model itself, that step is a model call like any other and you pay for it on the runs where it happens, because you put it there, not because the automation needs a model to know what to do.

Who builds Whenever?

Wix.com Ltd, the Nasdaq-listed company (ticker WIX). Whenever is one of its products, not a site built on the Wix website platform.

  • Wix