FLOWWATCH DUTY LOG · FORM FW-1

Your clients' automations fail at 3 AM.
Your clients never find out.

FlowWatch watches every client's n8n and Make workflows and explains each failure in plain English, with a suggested fix. At month-end it hands you the white-label report that proves what the retainer bought.

Claim a founding slot

$49/mo, locked for life. First 10 agencies. No card now: join the list and I set up your instance personally.

WHILE YOU WERE OUT FORM FW-2 · INCIDENT

TO: On-call (you)

CLIENT: Cedarview Dental

TIME: 03:12

  • Invoice sync → accounting: 14 invoices stuck since 02:10
  • Cause: their API added a required field (tax_region)
  • Suggested fix sent to your Discord
  • Client noticed
Demonstration data from FlowWatch's built-in simulator, with a fictional client.

One log across every client

n8n instances are polled every five minutes by default; error-trigger webhooks land instantly; Make error handlers point at the same URL. Every failure reaches your Discord already explained: cause first, fix attached, repeats throttled.

DUTY LOG — NIGHT OF JUL 23 ALL CLIENTS

02:10 cedarview · invoice-sync FAILED ×1

02:11 cause logged: accounting API now requires tax_region; fix suggested

02:15 alert → #cedarview-ops on Discord delivered

03:40 harborline · lead-router TIMEOUTS ×21 (provider outage)

04:02 harborline · lead-router succeeded AUTO-RECOVERED

06:40 cedarview · invoice-sync succeeded after your fix RESOLVED

07:00 morning digest → Discord: 8 clients · 4,211 executions · 2 incidents, 0 unresolved

Demonstration data: every line is a real FlowWatch event type.

Month-end, tear off the copy

Everything FlowWatch logged becomes a monthly report for each client, white-labeled so it reads as your work, because it is. When the client asks what the retainer bought, you attach this to the invoice.

Demonstration report rendered from simulator data. FlowWatch's name appears nowhere on real client reports.

Open the full sample report — the exact page a client receives →

Three entries and you're covered

  1. CONNECT FW-4a

    Paste a client's n8n API key, or drop FlowWatch's error-webhook URL into an Error Trigger node. Make error handlers use the same URL. A client is connected in minutes, not sprints.

  2. SLEEP FW-4b

    Failures reach your Discord in plain English with the likely cause and a suggested fix. Repeats are throttled so a flapping workflow can't bury you. Incidents auto-resolve on the next successful run.

  3. TEAR OFF FW-4c

    Each client's report lives at its own private link, rebuilt monthly from the log. Your logo, your name. Attach it to the invoice and the "what am I paying for?" conversation is already answered.

Read before signing

What you get: your own isolated FlowWatch instance, one agency per deployment, your data sharing nothing with anyone. n8n and Make coverage. Discord alerts, a daily digest at your hour, white-label monthly reports for every client. No per-client pricing.

What it doesn't do yet: email alerts (Discord only), automatic AI-spend tracking (you enter it monthly if you want it on reports), or re-running failed executions; n8n already does that, and FlowWatch tells you which ones and why.

Why not build it yourself: you could wire the alert half in n8n this afternoon, for one client. The cross-client log and the monthly report every client gets are the part nobody builds for themselves.

And plainly: every client and workflow number on this page comes from the product's built-in simulator, labeled as such. No invented customers, no fake logos. You'd catch me anyway — checking claims is the whole product.

The founding rate

FOUNDING RATE FIRST 10 AGENCIES

$49/month — locked for life

  • Your own isolated instance, set up personally by the founder
  • All clients and workflows included — no per-client pricing
  • White-label monthly reports, Discord alerts, daily digest
  • A direct line to me; founding requests shape the roadmap

Standard pricing will be higher. The founding rate never changes for founding agencies.