aws-health-monitor_skill

This skill monitors AWS Health Dashboard for active incidents and notifies via configurable channels, helping you stay informed of outages and updates.

Python

2.5k

GitHub Stars

2

Bundled Files

2 months ago

Catalog Refreshed

3 months ago

First Indexed

Readme & install

Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.

Installation

Preview and clipboard use veilstrat where the catalogue uses aiagentskills.

npx veilstrat add skill openclaw/skills --skill aws-health-monitor

_meta.json296 B
SKILL.md3.6 KB

Overview

This skill monitors the AWS Health Dashboard for active incidents and pushes notifications to configurable channels like Feishu, Telegram, Slack, and Discord. It watches specified AWS regions and services, extracts concise root-cause hints from event text, and sends updates when incidents appear, change, or resolve. Supports HTTP/HTTPS proxy for restricted networks.

How this skill works

The monitor polls AWS Health public events JSON and keeps a local state to detect new, updated, or cleared incidents. It filters by region and service white-lists, applies ignore rules (by event ARN or service name), and extracts root-cause keywords from event messages using pattern matching. When a change is detected, it formats a short notification and delivers it via the configured channel and target identifier.

When to use it

You want real-time alerts for AWS outages affecting your regions or services.
You need consolidated incident summaries delivered to team chat platforms.
You must monitor only specific regions (e.g., eu-central-1, ap-northeast-1, me-central-1).
You operate behind a corporate proxy and require HTTP(S) proxy support.
You want to auto-ignore noise by service or by specific event ARN.

Best practices

Set AWS_HEALTH_WATCH_REGIONS and AWS_HEALTH_WATCH_SERVICES to limit noise and API usage.
Populate the ignore JSON with known noisy ARNs or non-actionable services.
Run the script on a cron job every 3–10 minutes to balance freshness and rate limits.
Use a dedicated notify target/channel per team to route relevant incidents correctly.
Enable logging and retain the local state file for consistent deduplication and delta detection.

Example use cases

Notify SREs in Slack when EC2 or RDS incidents occur in Frankfurt and Tokyo.
Send Feishu alerts for any service impact in the UAE (me-central-1) region.
Filter to only monitor a subset of services to reduce alert fatigue during platform maintenance.
Operate the monitor from a bastion host using HTTP_PROXY/HTTPS_PROXY in a locked network.
Auto-skip recurring informational events by adding their ARNs to the ignore list.

FAQ

The monitor covers global events and region-specific events including eu-central-1, ap-northeast-1, and me-central-1; you can also leave region whitelist empty to watch all regions provided by the public feed.

How are root causes identified?

Root causes are inferred by regex matching common keywords in event messages (network, power, hardware, DNS, certificate, capacity, etc.); multiple matches are joined, and unmatched events show a default "no clear root cause" message.

How do I prevent repeated alerts for the same incident?

The skill maintains a local state file to detect and suppress duplicates; it sends notifications only for new events, message updates, or when an event clears.