n8n, Zapier and Make

Automation flows don't need to read instructions. They just send the report. You add two steps: one at the end of the flow for good runs, and one on the error branch for bad ones. The error branch is what turns a silent failure into a Failed run on your list.

1. Add the agent

In the dashboard, add the agent with its schedule and what success looks like. Copy the link Mirtel shows you; she shows it only once. Your link is:

https://YOUR-MIRTEL-HOST/r/mrt_agent_…

2. Add an HTTP step at the end of the flow

Add an HTTP Request node (n8n), a Webhooks by Zapier POST action (Zapier) or an HTTP module (Make), set up like this:

The body, for a good run:

{"status":"done","items":5,"output":"5 invoices chased"}

Map items to the number of things the flow produced, and output to a short summary or a link to the result. If the flow ran but found nothing, send "status":"empty" instead.

No Authorization header is needed. The link is the key.

3. Add another on the error branch

Add the same HTTP step to the flow's error path (an error workflow in n8n, an error handler route in Make, or an error path in Zapier where your plan has one), with:

{"status":"failed","error":"what went wrong"}

Map error to the error message if your tool gives you one.

If a step needs a human decision, you can send "status":"needs_you" with what's needed in output.

Check it worked

Run the flow once by hand. The run should appear in the ledger. Then break something on purpose and check that it shows as Failed, not Missed.

Keep the link safe

Anyone with the link can report as this agent. Store it in your tool's credentials or variables if it has them. If it leaks, rotate the agent's key in the dashboard; the old link stops working at once.