Skip to content
TakoAPI
Back to skills

pager-triage

by tkuehnl

Coding Agents & IDEs
—Not yet rated
Your rating

Brief

AI-powered incident triage for PagerDuty.

Description

What This Skill Does

Connects to PagerDuty's REST API to list active incidents, view timelines and alert correlations, check on-call schedules, and take actions like acknowledge, resolve, and annotate. Read-only by default; write operations require an explicit --confirm flag.

Handles the full triage loop, from listing incidents to acknowledging and annotating, without leaving the agent, and enforces confirmation gates on all write operations to prevent accidental changes.

When to Use It

  • Checking what incidents are currently firing during an outage
  • Finding who is on call for a specific service right now
  • Acknowledging a triggered incident to stop further escalation
  • Adding a root cause note to an incident's timeline
  • Reviewing recent incident history for a service over the past 7 or 30 days

Example Workflow

INPUT

User asks: What's currently firing on PagerDuty?

AGENT
  1. 1

    Calls pd_incidents to retrieve all triggered and acknowledged incidents

  2. 2

    Sorts results by urgency (high first) and duration (oldest first)

  3. 3

    Calls pd_incident_detail on the highest urgency incident to get timeline and alert correlation

  4. 4

    Returns a summary of all active incidents with detailed context on the most critical one

OUTPUT

List of 3 active incidents with the highest-urgency incident's full timeline, trigger source, escalation history, and recommended next steps

Requirements

  • PAGERDUTY_API_KEY environment variable (PagerDuty REST API v2 token)
  • PAGERDUTY_EMAIL environment variable (required only for write operations: ack, resolve, note)

Installation

clawhub install tkuehnl/pager-triage

Run the CLI command, or paste the prompt into Claude Code, Cursor, Codex or whichever coding agent you use.

Links