---
title: The API refuses the monitor settings the dashboard refuses
description: Through the API, MCP and A2A you could save settings that every check then failed on, or that were quietly ignored: a private or hostname DNS resolver, an…
canonical: https://watchfor.io/changelog/2026-11-22-api-refuses-settings-the-dashboard-refuses
---

[All updates](/changelog#2026-11-22-api-refuses-settings-the-dashboard-refuses) 🐛 Fixed API Monitors Nov 22, 2026 · v 13.8.9

# The API refuses the monitor settings the dashboard refuses

Through the API, MCP and A2A you could save settings that every check then failed on, or that were quietly ignored: a private or hostname DNS resolver, an https:// WebSocket target, an invalid regular expression, a UDP check with nothing to send, an MTR count of 100, a confirmation threshold of 101, an interval of years. They were stored and the monitor went down or never alerted. Every write path now checks the same rules and answers 400 naming the field. Monitors you already have are unchanged.

## Now refused with a 400 that names the field

- DNS and blacklist resolver : must be a public IP address, optionally with a port ( 1.1.1.1 , 8.8.8.8:53 ). Hostnames and private addresses were refused by the check every time.

- Blacklist : RBL zones that are not DNS zones, and private IP addresses as the target (they are never on a public blacklist, so every lookup read as clean).

- WebSocket target with http:// or https:// : the answer names the ws:// / wss:// URL to use.

- Breach target with a path: breaches are looked up per domain, the path was never checked.

- Regular expressions (TCP/UDP response match, WebSocket pattern, DNS regex match, HTTP body match) are compiled when you save, in the RE2 syntax the check uses.

- UDP without anything to send.

- Counts : ping, MTR and NTP counts above 10.

- Alerting : confirmation threshold outside 1–100, duration window outside 0–86,400 seconds, repeat reminders that are not whole hours up to 24.

- Interval above 24 hours (an interval past 2,147,483,647 seconds used to end in a bare 500).

- Basic and Digest auth without a username on API monitors: the check used to go out unauthenticated. The dashboard now refuses it too.

- SFTP private key that is not a key, or that has a passphrase: every check failed before the password was tried.

- Core Web Vitals device : fixed after creation over the API too, as the dashboard says.

Nothing changes for monitors you already have: none stored a value these rules refuse.

Share this update

---

Canonical page: https://watchfor.io/changelog/2026-11-22-api-refuses-settings-the-dashboard-refuses · Site guide: https://watchfor.io/llms.txt
