For developers
The monitoring tool you'd have built - if you had the week to spare.
Sub-minute checks on real Chromium. JSON diff for API monitors. A clean REST API for everything. And an attribution engine that tells you which commit caused the regression before you start grepping git log.
The problem
Why this keeps showing up in your week.
Post-deploy regressions are nobody's job
Sentry catches errors. Datadog catches latency. Nobody catches the silent regression where the hero button rename breaks the marketing team's tracking. By the time someone notices, it's a 3pm "why did this change?" thread.
API consumers break without warning
Third-party APIs change response shapes, deprecate fields, and shift rate limits with no notice. If your integration suddenly returns nulls, your monitoring should tell you before your users do.
Every incident starts with "what changed?"
You know the drill - grep git log, crack open the deploy timeline, page the last person who touched the repo. 30 minutes of forensics for a 30-second revert.
Why OnChange
The specific parts that make this work.
Full REST API, not an afterthought
Create, update, delete, and query monitors programmatically. Pull change history, diff data, and alert logs as JSON. Rate-limited at 120 req/min per user. See /api-docs - the catalog reflects real endpoints, not marketing copy.
Sub-minute checks, real Playwright
Monitors run in a real headless Chromium, so JavaScript-rendered SPAs work without custom config. Business plans check every 30s; Enterprise drops to 10s. Works for flash-sale detection, API regression catching, and everything in between.
Attribution on every change
Configure github_repo + a read-only token on any monitor. When OnChange detects a change, it pulls commits from the 15-minute window before, passes the diff and commit list to Claude, and returns a verdict like "Likely commit abc1234 - matches the renamed header link."
JSON diff for API monitors
Point a monitor at a REST or GraphQL endpoint with custom headers, methods, and request bodies. Diffs highlight which fields changed, were added, or were removed - clean enough to paste into a post-mortem.
Webhooks + Slack + Discord + email
Route alerts to any channel, per monitor. Webhook payloads include the full diff as structured JSON with signed HMAC headers so your intake endpoint can trust the source.
Every feature on every plan
No surprise "that's a Pro feature" walls. Visual diffs, API monitoring, AI summaries, REST API, attribution, WCAG scans - all on Free. Paid plans differ only in monitor count and check frequency.
Wire it into CI/CD
Create a monitor from a deploy script, pass a metadata tag so alerts route to the right Slack channel, and delete the monitor on teardown. OnChange's REST API is straightforward - three POSTs to stand up a full page + API monitor suite per environment.
API docsCatch your own deploys before your users do
Hook attribution to your GitHub org and every staging/prod change gets matched to the commit that caused it. When something regresses, the alert includes the suspected SHA, so you can revert in a minute instead of debugging for an hour.
How attribution worksDeploy with an extra set of eyes.
Start free, script your first monitor from the API in two minutes, and watch your next deploy tell you what changed.