alert-router

What problem does it solve?

This Skill helps you manage real-time anomaly alerts by separating true fraud signals from noisy event streams, then routing the important ones to storage and notifications with minimal duplication.

Core Features & Use Cases

• Severity bucketing: Converts anomaly scores into LOW, MEDIUM, HIGH, or CRITICAL alerts so responders can prioritize quickly.
• Redis deduplication: Prevents repeated alerts for the same user and severity using TTL-based keys.
• Database and Slack fanout: Writes de-duplicated alerts to TimescaleDB and optionally posts high-severity events to Slack.
• Use Case: A fraud monitoring pipeline can feed scored transactions into this Skill to suppress repeats, persist alert history, and notify the operations team only when needed.

Quick Start

Use the alert-router skill to explain how anomaly scores are bucketed, deduplicated in Redis, written to TimescaleDB, and sent to Slack.