Integrate Sentry for Error Monitoring

The Sentry MCP server integrates Sentry.io error tracking and application monitoring into Claude's workflow. It enables AI-assisted debugging by exposing tools to list organizations and projects, view unresolved issues, analyze error stack traces, examine event details and breadcrumbs, and resolve issues directly through conversational commands.

Use this connector when debugging production errors that require context from Sentry's error tracking data, prioritizing bug fixes based on error frequency and impact, investigating user-reported issues with full event history, or monitoring application health across multiple projects. Use it when you need to quickly triage issues and understand error patterns without leaving your AI assistant.

Do NOT use this connector if you need to modify Sentry project settings, configure alerts, or manage team permissions - it provides read access to issues and events plus basic issue resolution, not administrative capabilities. Avoid it for real-time alerting workflows where Sentry's native notification system is more appropriate.

You provide a Sentry auth token with project:read, issue:read, event:read, and org:read scopes. For queries, you supply organization slugs, project slugs, issue IDs, and optional query filters.

You receive structured data including organization lists, project lists, issue details with stack traces, event lists with breadcrumbs, error frequency metrics, issue status and assignments, and confirmation of resolution actions.

This MCP server integrates exclusively with Sentry.io, the error tracking and application monitoring platform. It requires a valid Sentry account and auth token. Install via uvx (recommended) or pip:

uvx mcp-server-sentry

Configure in Claude Code settings:

{
  "mcpServers": {
    "sentry": {
      "command": "uvx",
      "args": ["mcp-server-sentry"],
      "env": {
        "SENTRY_AUTH_TOKEN": "your_auth_token_here"
      }
    }
  }
}

The connector provides tools including list_organizations(), list_projects(organization_slug), list_issues(organization_slug, project_slug, query?), get_issue(issue_id), get_issue_events(issue_id), and resolve_issue(issue_id).

Software engineers and DevOps teams who use Sentry for error tracking and want AI-assisted debugging workflows. Ideal for developers who triage production issues, investigate error patterns, and need quick access to stack traces and event context without switching between tools. Engineering managers can use it to understand error impact and prioritize fixes based on frequency data. On-call engineers benefit from rapid issue analysis during incident response.