incident-alert-tickets

Look up and update Linear incident-alert tickets for Datadog monitors and on-call pages.

34.0k|3.7k|Updated May 18, 2023
One-click install
npx skills add https://github.com/langfuse/langfuse --skill incident-alert-tickets
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: incident-alert-tickets
Source: https://github.com/langfuse/langfuse/tree/main/.agents/skills/incident-alert-tickets
Command: npx skills add https://github.com/langfuse/langfuse --skill incident-alert-tickets

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

On-call debugging knowledge gets lost after each incident, forcing engineers to re-diagnose the same alert root causes from scratch. This Skill turns Linear tickets into a searchable per-monitor knowledge base so past root causes and fixes are found before re-investigating.

Core Features & Use Cases

  • Alert Lookup: Match a Datadog monitor ID, incident.io alert, or on-call page to its Linear ticket via the incident-alert label and read documented root causes.
  • Cause Classification: Compare current evidence against each cause section's recognition signals and classify as known cause, new cause, or missing ticket.
  • Human-Gated Write-Back: Propose appending dated cause sections or creating new tickets, but only write after explicit human approval.
  • Use Case: You get paged for a Datadog monitor. Before diving into logs, look up the monitor's ticket, find a matching known cause with a documented fix, and resolve the incident in minutes instead of hours.

Quick Start

Look up the incident-alert ticket for Datadog monitor 12345 and tell me if any documented root cause matches the current error spike.

Frequently Asked Questions about incident-alert-tickets

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I find past root causes for a Datadog monitor alert?

List Linear issues carrying the incident-alert label, then match on the monitor ID stored in the ticket's alert header, falling back to monitor title and environment. Read the matched ticket's dated cause sections and comments for documented recognition signals and fixes.

How do I document a new incident root cause in Linear?

Propose a dated cause section with Recognize it, How urgent, and Fix fields, presented in an approval table first. Only after explicit human approval is the section appended above the ticket's Your cause is not listed trailer.

Can this skill create or update Linear tickets automatically?

No, all writes are gated on explicit human approval. The skill presents a proposal table with classification, proposed action, and draft content, then waits for the human to select rows before creating or appending anything.

What happens when multiple alerts fire at the same time?

Each alert identity gets its own lookup, comparison, and classification because every monitor has its own ticket. If one root cause explains several alerts, the full cause section goes on the monitor closest to the cause, with short linking sections proposed on the others.

When should I not use incident-alert ticket lookup?

Skip it when there is no alert identity, such as a customer report or general code question with no named Datadog monitor, incident.io alert, or on-call page. The presence of a named alert is the mandatory trigger condition.