For a couple of personal services with a linear trigger-and-action, this is genuinely one of the simpler builds here — direct API calls in a small script, not a general routing platform.
Build it
Effort: an eveningRunning cost: $0/mo to start
Why CanICodeThis says Build it
A single trigger (a new item appears, a time of day arrives) firing one action is a bounded, well-documented pattern
For a couple of personal services you already use, direct API calls are simpler than a general routing platform
Your automations become a small script in a repo: free, versioned, and easy to understand
The parts that'll cause problems
A pre-built connector catalog covering hundreds of services represents years of integration maintenance
Reliable, always-on scheduling for many users' automations needs real infrastructure beyond a personal script
Visual branching logic (if this, do A, else do B) is a small workflow engine, a real, separate piece of engineering from a linear script
How I'd build this
runtime
A small scheduled Cloudflare Worker or serverless function per automation
hosting
Cloudflare Workers
MVP scope
A trigger (webhook or scheduled check) firing one action
Direct API calls to one or two services you already use
Basic logging of what ran and when
Postpone to v2
A connector catalog for more services
Visual branching logic
A web UI instead of editing code directly
What it actually costs to build
AI agent — with a subscription (Claude Pro/Max, Cursor, etc.)
$0 marginal
AI agent — pay-per-use API, no subscription
$16–$27 one-time
Hosting, once it's running
$0/mo (free tier)
Domain name, if you want your own
~$12/yr
Existing tools solving a similar problem typically run $5–$30/mo — see how they compare below.
The build prompt
Genuinely one of the simpler builds here — even simpler in scope than this register's Zapier and Make entries:
1. A trigger (webhook or scheduled check) firing one action — a bounded, well-documented pattern.
2. Direct API calls to one or two services you already use, with credentials from environment variables.
3. Basic logging of what ran and when.
A small serverless function per automation is genuinely sufficient for personal use — your automations become code in a repo, versioned and free at any volume. Out of scope: a connector catalog for more services, which represents years of integration maintenance, and visual branching logic, a small, separate workflow engine.
