返回 MCP 目录
public公开dns本地运行

Sentry模型上下文协议服务器

一种模型上下文协议服务器,允许AI助手与Sentry API交互以检索和分析错误数据、管理项目和监控应用程序性能。

article

README

Sentry MCP Server

A Model Context Protocol (MCP) server for interacting with Sentry. This MCP server provides tools to interact with the Sentry API, allowing AI assistants to retrieve and analyze error data, manage projects, and monitor application performance.

Requirements

  • Node.js (v14 or higher)
  • npm or yarn
  • Sentry account with API access
  • Sentry authentication token with appropriate permissions

Setup

  1. Install dependencies:
    npm install
    

Using this within an IDE

This MCP has been verified to work against Codeium Windsurf.

Cursor is currently having issues with its MCP implementation; and this tool is not yet fully functional.

Using with Claude

To use this MCP server with Claude, add the following configuration to your Claude settings:

{
    "mcpServers": {
        "sentry": {
            "command": "npx",
            "args": ["ts-node", "/Users/<your-user-directory>/sentry-mcp-ts/index.ts"],
            "env": {
                "SENTRY_AUTH": "<YOUR_AUTH_TOKEN>"
            }
        }
    }
}
  • Update with your directory path in the args field.
  • Replace <YOUR_AUTH_TOKEN> with your Sentry authentication token.

Available Tools

list_projects

Lists all accessible Sentry projects for a given organization.

Parameters:

  • organization_slug (string, required): The slug of the organization to list projects from
  • view (string, optional): View type, either "summary" or "detailed" (default: "detailed")
  • format (string, optional): Output format, either "plain" or "markdown" (default: "markdown")

resolve_short_id

Retrieves details about an issue using its short ID.

Parameters:

  • organization_slug (string, required): The slug of the organization the issue belongs to
  • short_id (string, required): The short ID of the issue to resolve (e.g., PROJECT-123)
  • format (string, optional): Output format, either "plain" or "markdown" (default: "markdown")

get_sentry_event

Retrieves and analyzes a specific Sentry event from an issue.

Parameters:

  • issue_id_or_url (string, required): Either a full Sentry issue URL or just the numeric issue ID
  • event_id (string, required): The specific event ID to retrieve
  • view (string, optional): View type, either "summary" or "detailed" (default: "detailed")
  • format (string, optional): Output format, either "plain" or "markdown" (default: "markdown")

list_error_events_in_project

Lists error events from a specific Sentry project.

Parameters:

  • organization_slug (string, required): The slug of the organization the project belongs to
  • project_slug (string, required): The slug of the project to list events from
  • view (string, optional): View type, either "summary" or "detailed" (default: "detailed")
  • format (string, optional): Output format, either "plain" or "markdown" (default: "markdown")

create_project

Creates a new project in Sentry and retrieves its client keys.

Parameters:

  • organization_slug (string, required): The slug of the organization to create the project in
  • team_slug (string, required): The slug of the team to assign the project to
  • name (string, required): The name of the new project
  • platform (string, optional): The platform for the new project
  • view (string, optional): View type, either "summary" or "detailed" (default: "detailed")
  • format (string, optional): Output format, either "plain" or "markdown" (default: "markdown")

list_project_issues

Lists issues from a specific Sentry project.

Parameters:

  • organization_slug (string, required): The slug of the organization the project belongs to
  • project_slug (string, required): The slug of the project to list issues from
  • view (string, optional): View type, either "summary" or "detailed" (default: "detailed")
  • format (string, optional): Output format, either "plain" or "markdown" (default: "markdown")

list_issue_events

Lists events for a specific Sentry issue.

Parameters:

  • organization_slug (string, required): The slug of the organization the issue belongs to
  • issue_id (string, required): The ID of the issue to list events from
  • view (string, optional): View type, either "summary" or "detailed" (default: "detailed")
  • format (string, optional): Output format, either "plain" or "markdown" (default: "markdown")

get_sentry_issue

Retrieves and analyzes a Sentry issue.

Parameters:

  • issue_id_or_url (string, required): Either a full Sentry issue URL or just the numeric issue ID
  • view (string, optional): View type, either "summary" or "detailed" (default: "detailed")
  • format (string, optional): Output format, either "plain" or "markdown" (default: "markdown")

list_organization_replays

Lists replays from a specific Sentry organization.

Parameters:

  • organization_slug (string, required): The slug of the organization to list replays from
  • project_ids (string[], optional): List of project IDs to filter replays by
  • environment (string, optional): Environment to filter replays by
  • stats_period (string, optional): Time period for stats (e.g., "24h", "7d")
  • start (string, optional): Start date for filtering replays
  • end (string, optional): End date for filtering replays
  • sort (string, optional): Field to sort replays by
  • query (string, optional): Search query to filter replays
  • per_page (number, optional): Number of replays per page
  • cursor (string, optional): Cursor for pagination
  • view (string, optional): View type, either "summary" or "detailed" (default: "detailed")
  • format (string, optional): Output format, either "plain" or "markdown" (default: "markdown")

Running the Server

npx ts-node index.ts

Authentication

This tool requires a Sentry authentication token with appropriate permissions to access the Sentry API. You can generate a token in your Sentry account settings under "API Keys".

Error Handling

The server includes comprehensive error handling for:

  • Missing authentication token
  • API request failures
  • Invalid parameters
  • Network errors

All errors are logged to the console for debugging.

help

运行方式说明

cloud

托管运行

托管运行通常表示这个 MCP Server 由服务方环境承载,用户一般按页面提供的连接方式或授权流程接入,不需要在本地长期启动一个 MCP 进程

  1. 打开服务方连接页
  2. 完成授权或复制端点
  3. 在 MCP 客户端中连接
terminal

本地运行 / 其它方式

本地运行通常需要用户在自己的电脑或服务器上安装依赖,把 server_config 复制到 MCP 客户端,并按 env_schema 补齐环境变量、密钥或其它配置

  1. 复制 server_config
  2. 安装所需依赖
  3. 补齐环境变量后重启客户端