API
Connect Make¶
Read your remake results in your Make scenario, start runs from it, and send your events into it.
Before you start¶
- A Make account: you build your scenario there.
- The account owner's sign-in: you need it, because viewer seats can't create keys or channels.
- A free key slot: the Make tile creates an API key for you, and your account holds 5 live keys at most.
- To start runs: the brand needs a schedule and your wallet needs credit.
Connect Make¶
- Open Integrations.
- On the Connections tab, find Make under Automations and click Connect.
- In the Connect Make dialog, click Copy and store your key somewhere safe.
- Click Next. Step 2 lists the event steps, covered below.
- Click Next again. On step 3, click the blueprint link to download the file.
- Click Done.
The key is shown once
Your key is named Make connector under API keys on Your AI. It is shown once, so you can't see it again.
Send events into Make¶
- In Make, start your scenario and add the Webhooks module.
- Copy the module's address.
- In remake, open the Alerts tab on Integrations.
- Under Add a channel, keep Webhook (HTTPS) and paste the address.
- Click Add channel.
- Open the new row's menu and click Send test. Make learns the payload shape from it.
Your scenario then hears the same events as a Zap. The list is in What arrives.
Read your data with the blueprint¶
- In Make, open a new scenario, then ⋯ › Import Blueprint, and pick the file you downloaded.
- In the HTTP module, swap
hf_YOUR_KEYfor your key. KeepBearerand the space after it. - Run the scenario once. The module returns your workspaces and the brands in each.
To read something else, you change only the module's address. Every read is a GET with the same header.
| To read | Address after https://app.tryremake.com |
|---|---|
| Workspaces and their brands | /v1/workspaces |
| A brand's headline numbers | /v1/brands/{brand}/dashboard |
| Visibility over the latest 24 runs | /v1/brands/{brand}/visibility |
| A brand's runs | /v1/brands/{brand}/runs |
| One run's answers | /v1/brands/{brand}/runs/{run_id}/answers |
| Wallet balance and spend | /v1/wallet |
{brand} is the brand's website domain, such as samsung.com. The reference lists every address.
Start a run from a scenario¶
Make has no ready-made module for this. You build it from HTTP modules: one asks for a quote, the next confirms it.
-
Add an HTTP module that sends a POST, with the same Authorization header and no body, to this address:
https://app.tryremake.com/v1/brands/samsung.com/prompt-spend/run-now -
Read the answer you get back. It holds
price_mills,quote_tokenandwould_refuse. Nothing is charged. - Continue only when
would_refuseis empty. Otherwise it names the reason. -
Add a second POST to the same address, with this JSON body:
{"confirm": true, "quote_token": "<quote_token from step 2>", "request_id": "run-2026-10-07"} -
Read
run_idfrom its answer. The status isqueued. - Every minute or two, GET
/v1/brands/{brand}/runs/{run_id}untilstatusisdoneorfailed.
The confirm spends credit
The second POST starts a real run at the quoted price and charges your wallet. It is the brand's Run now, so the next scheduled run moves one cycle out.
A full run typically takes 3 to 6 minutes. Prices are in mills, and 1,000 mills make one dollar. Each ask is charged as it runs, and asks that fail on our side are refunded.
A quote lasts 30 minutes and works once. Give each scenario a fresh request_id for every run, such as the date. A retry with the same id within 24 hours returns the first answer instead of a second run.
Check it worked¶
- The Make tile reads Connected, and its button reads Manage.
- Under API keys on Your AI, the
Make connectorkey shows when it was last used. - The blueprint's module returns your workspaces with no error.
- After Send test, the Delivery log shows a
testrow with the status sent.
FAQ¶
I closed the dialog before copying the key. What now?¶
Revoke the Make connector key under API keys on Your AI. You then get Connect on the Make tile again, with a fresh key.
Why did my confirm return quote_changed?¶
The price moved, or the quote expired or was used already. The answer carries a fresh quote. You confirm again with its quote_token.
Why was the run refused?¶
The refusal names the reason. The schedule may be paused or missing, or the month's budget reached. Your wallet may not cover the run, a run may already be going, or 3 runs may be queued or running.
How many calls can a scenario make?¶
120 a minute for your whole account, shared with everything else that uses your keys and connected apps. Past that, a call answers 429 with the seconds to wait. Calls are free.