Skip to content
modsignal

modsignal documentation

modsignal watches public web pages and alerts on meaningful change, described in plain language. These pages document exactly what ships today — no roadmap items.

Quick start

  1. Create a free account — 3 monitors with daily checks, no credit card.
  2. Create a monitor: paste the URL, then describe the change you care about in a sentence — “Tell me if a price or plan limit changes. Ignore copy edits.” Or start from a template.
  3. Pick a check tier and interval. The first run captures a baseline; after that, you only hear from us when your sentence comes true.
  4. Read the alert: one sentence, the before, the after, and a confidence score — in your inbox, Slack channel, webhook receiver, or RSS reader. Alert channels →

Monitors

Creating monitors, writing prompts that stay quiet, schedules, templates, pausing, and what happens when a page keeps failing.

Check tiers

Exact diff, semantic AI, and browser agent — how each decides what changed, what evidence each produces, and how to pick.

Alerts & evidence

The anatomy of an alert, plus every delivery channel: email, Slack, signed webhooks with heartbeats, and the RSS feed.

Teams & billing

Roles and permissions, plan limits, what upgrades and downgrades actually do, and how cancellation works.

API reference

The REST API: authentication, rate limits, endpoints, webhook payloads, and signature verification.

How modsignal thinks

Two ideas run through everything documented here. First: you describe the change, not the page region — there are no CSS selectors to maintain, and noise gets filtered by meaning, not by pixel masks. Second: every alert carries its own proof — the sentence, the before/after evidence, a confidence score, and on the browser-agent tier a screenshot of the page as checked. You should never have to open the page to verify what we told you.