Sentry

This MCP server lets you interact with Sentry programmatically, enabling AI assistants to fetch error data, inspect issues, manage projects, and monitor performance through a streamlined MCP API.

How to use

You interact with the Sentry MCP Server through an MCP client that can send function calls such as listing projects, resolving issues, retrieving events, and listing replays. Each tool corresponds to a specific operation against the Sentry API. Use the available tools to retrieve data, analyze issues, and monitor project health from your AI workflows or automation scripts.

How to install

Prerequisites include Node.js v14 or higher, npm or yarn, and a Sentry account with API access. You also need a Sentry authentication token with the required permissions.

npm install

Install dependencies locally to prepare the MCP server for running.

Start the MCP server with the runtime command shown for the provided configuration example.

Configuration and usage notes

To run the MCP server locally and expose it to your MCP client, use the following runtime configuration that specifies the command, arguments, and required environment variables.

{
  "mcpServers": {
    "sentry": {
      "type": "stdio",
      "name": "sentry",
      "command": "npx",
      "args": ["ts-node", "/Users/<your-user-directory>/mcp-sentry-ts/index.ts"],
      "env": {
        "SENTRY_AUTH": "<YOUR_AUTH_TOKEN>"
      }
    }
  }
}

Authentication and security

The server requires a Sentry authentication token with appropriate permissions. Generate the token in your Sentry account settings under API Keys and provide it to the MCP as SENTRY_AUTH in the environment.

Error handling and troubleshooting

The server includes robust error handling for missing authentication tokens, API request failures, invalid parameters, and network errors. All errors are logged to the console to facilitate debugging.

Additional notes

This MCP server supports a suite of tools to interact with Sentry data, including project management, issue resolution, event retrieval, and replay listing. Use these tools to build automated workflows that analyze errors, correlate issues across projects, and monitor application health over time.

Available tools

list_projects

Lists all accessible Sentry projects for a given organization. Parameters include organization_slug, view, and format.

resolve_short_id

Retrieves details about an issue using its short ID, given organization_slug and short_id.

get_sentry_event

Retrieves and analyzes a specific Sentry event from an issue using issue_id_or_url and event_id.

list_error_events_in_project

Lists error events from a specific Sentry project within an organization, with optional filters for view and format.

create_project

Creates a new project in Sentry and retrieves its client keys, requiring organization_slug, team_slug, name, and optional platform.

list_project_issues

Lists issues from a specific Sentry project within an organization.

list_issue_events

Lists events for a specific Sentry issue.

get_sentry_issue

Retrieves and analyzes a Sentry issue by ID or URL.

list_organization_replays

Lists replays from a specific Sentry organization with optional filters for environment, stats, and pagination.