PagerDuty MCP Server

Name: PagerDuty MCP Server
Author: wpfleger96

@wpfleger96

关于 PagerDuty MCP Server

MCP server exposing PagerDuty incidents, services, schedules, and on-calls to AI agents — read, acknowledge, resolve, with OAuth and API key auth

基本信息

分类

AI 与智能体

许可证

MIT

运行时

python

传输方式

stdio

发布者

wpfleger96

配置

使用下面的配置,将此服务器添加到你的 MCP 客户端。

{
  "mcpServers": {
    "pagerduty-mcp-server": {
      "command": "uv",
      "args": [
        "sync"
      ]
    }
  }
}

工具

未检测到工具

工具是从 README 中自动提取的。维护者可以在 ## Tools 标题下列出工具,即可填充这部分内容。

概览

What is PagerDuty MCP Server?

It is a server that exposes PagerDuty API functionality to LLMs, designed for programmatic use with structured inputs and outputs. It allows LLMs to perform operations on PagerDuty resources such as incidents, services, teams, and users.

How to use PagerDuty MCP Server?

Install via uvx pagerduty-mcp-server and configure authentication using a PagerDuty API token, OAuth 2.0 PKCE, or the X-PagerDuty-Token HTTP header. Use with Claude/Cursor by adding the MCP configuration, or run as a standalone server with uv run pagerduty-mcp-server.

Key features of PagerDuty MCP Server

Exposes PagerDuty API tools for LLMs
Read and write operations on incidents, services, teams, users
Optional include parameter to reduce token usage
Consistent response format with metadata and optional error
Automatic pagination and rate limit handling
Supports API token, OAuth, and per-request token authentication

Use cases of PagerDuty MCP Server

Query and manage incidents assigned to you
Check upcoming on-call schedules for your team
Retrieve details about services, teams, and users
Acknowledge or resolve incidents and add notes
Filter results by user context to reduce irrelevant data

FAQ from PagerDuty MCP Server

What authentication methods are supported?

Three methods: the X-PagerDuty-Token HTTP header (per-request), the PAGERDUTY_API_TOKEN environment variable, and OAuth 2.0 PKCE for local interactive use. OAuth is opt-in and requires a client ID.

How does the `include` parameter work?

Most read tools accept an optional include parameter – a list of field names to return. When specified, only those fields are included in each response object, reducing token usage.

What is the response format for errors?

When an error occurs, the response includes an error object with message and code fields. The error object is only present on errors; otherwise, responses contain metadata and a resource list.