📣 Channels

kwatch can send the same clear alert to 56 providers. Pick the place your team already checks:

You want...	Try...
💬 Team chat	Slack, Discord, Microsoft Teams, Mattermost, Rocket.Chat
🚨 On-call pages	PagerDuty, Opsgenie, Zenduty, SIGNL4, Squadcast
📧 Email or SMS	Email, SendGrid, AWS SES, Twilio
🔗 Your own system	Custom Webhook, Ntfy, Gotify, n8n, Zapier
📋 Tickets	Jira, ClickUp, GitHub, GitLab, Gitea

🌟 Dedicated setup guides

The most common providers have a step-by-step page:

Channel	Config key	What you need
Slack	slack	Webhook URL or bot token
Discord	discord	Webhook URL
Microsoft Teams	teams	Webhook URL
Google Chat	googlechat	Webhook URL
Telegram	telegram	Bot token and chat ID
Email	email	SMTP server and password
PagerDuty	pagerduty	Integration key
Opsgenie	opsgenie	API key
Zenduty	zenduty	Integration key
Mattermost	mattermost	Webhook URL
Rocket.Chat	rocketchat	Webhook URL
Matrix	matrix	Homeserver, token, and room ID
DingTalk	dingtalk	Access token
FeiShu	feishu	Webhook URL
Custom Webhook	webhook	Endpoint URL

The complete provider reference lists every supported provider, every catalog field, its type, whether it is required, and whether it must come from a Secret. It is generated from the same provider catalog used by kwatch.sh.

The sidebar's All provider guides section gives each provider its own page and minimal configuration example.

🧩 Basic configuration

Put one or more providers under alert:. This example sends alerts to Slack:

alert:
  slack:
    webhook: "${file:/config/slack-webhook}"

You can configure several providers. kwatch sends each incident to all matching providers.

🎯 Send only the alerts you need

Routes filter alerts by namespace, severity, or reason. All conditions in one route must match:

alert:
  slack:
    webhook: "${file:/config/slack-webhook}"
    routes:
      - namespaces: ["production"]
        severities: ["high", "critical"]
        reasons: ["OOMKilled"]

If you do not add routes, the provider receives all alerts.

🔁 Delivery that can recover

The same delivery options work for every provider:

alert:
  slack:
    webhook: "${file:/config/slack-webhook}"
    retry:
      maxAttempts: 3
      delay: 5s
    fallback: pagerduty

• 🔁 Retry tries temporary failures again, such as timeouts and server errors.
• 🆘 Fallback sends to another configured provider when the first one fails.
• 📏 Compact reduces message size when your channel has strict limits.
• 🧵 Threaded mode keeps Slack updates together when using a bot token.

Invalid requests and bad credentials are not retried forever. They are reported so you can fix the provider configuration.

🔐 Keep credentials safe

The interactive manager stores credentials in a Kubernetes Secret and mounts them at /config. Use exact ${file:/absolute/path} references in the config. Plain credentials and environment substitutions in sensitive fields are rejected at startup. Read the configuration reference for details.

✅ Test your channel

1. Run kwatch lint to validate the configuration.
2. Run kwatch lint --check to test credentials for providers that support checks.
3. Enable healthCheck.diagnostics and call /test-alert if you need a real test message.

For installation help, start with Getting Started or Installation.