Skip to main content

Sentry MCP

The Sentry MCP Server brings Sentry’s full issue context into systems that support the Model Context Protocol (MCP). Connect via Cequence AI Gateway to access Sentry issues and errors, search for errors in specific files, query projects and organizations, manage DSNs, and invoke Seer for automated analysis and fix suggestions—all from your AI client.

1. Overview​

Sentry MCP is a remote, vendor-hosted MCP server provided by Sentry. You connect to it from Cequence AI Gateway; Sentry hosts and manages the server so you always get the latest functionality.

  • Server URL: https://mcp.sentry.dev/mcp
  • Transport: HTTP (Streamable HTTP with automatic SSE fallback)
  • Hosted by: Sentry

The server is production-ready; MCP as a technology is still evolving, so changes may occur. Feedback and issues can be reported on GitHub.

2. Supported authentication types​

TypeSupportedNotes
OAuth 2.0YesRecommended. Uses Dynamic Client Registration (DCR). Sign in with your existing Sentry organization; no pre-configured client ID needed.
Legacy remote (mcp-remote)YesFor clients that don’t support OAuth, use npx mcp-remote@latest https://mcp.sentry.dev/mcp; the gateway can expose this for you.
Local STDION/AFor self-hosted Sentry you can run the MCP server locally with a Sentry User Auth Token; see sentry-mcp.

When you add Sentry MCP in Cequence AI Gateway, authentication is typically OAuth 2.0 with DCR. You’ll be prompted to accept OAuth authorization, sign in via your Sentry organization, and grant the necessary permissions.

3. What can you do with this MCP server​

With the Sentry MCP server, you can:

  • Access Sentry issues and errors — Get full issue context, search, and analyze problems.
  • Search for errors in specific files — Find errors in a given file or across projects.
  • Query projects and organizations — List and query organization and project information.
  • Manage DSNs — List and create Data Source Names (DSNs) for projects.
  • Invoke Seer — Use Sentry’s AI agent (Seer) for root cause analysis and automated fix suggestions; get fix status and details.
  • Release and performance data — Query releases and access transaction/performance data.
  • Custom queries — Run complex searches across Sentry data.

The server provides 16+ tools and prompts for bringing Sentry context into LLM interactions.

4. Prerequisites​

Before adding Sentry MCP in Cequence AI Gateway, ensure you have:

  • Access to Cequence AI Gateway (e.g. beta.aigateway.cequence.ai)
  • A Sentry account and access to the organization(s) and projects you want to use
  • For OAuth authentication: an auth app with client credentials in your vendor account, unless the server supports Dynamic Client Registration (DCR). Sentry MCP supports DCR, so you typically do not need to create an OAuth app in Sentry.

If you join a new Sentry organization later, log out of the MCP integration in your client and log back in to refresh access.

5. Example workflows​

  • “Tell me about the issues in my project project-name.”
  • “Check Sentry for errors in components/UserProfile.tsx and propose solutions.”
  • “Diagnose issue PROJECT-123 and propose solutions.”
  • “Create a new project in Sentry for new-service-name and suggest local instrumentation.”
  • “Use Sentry’s Seer to analyze and propose a solution for issue PROJECT-456.”
  • “Show me the most recent releases for my organization.”
  • “Find all unresolved crashes in my React Native app.”

6. Connecting MCP server from Cequence AI Gateway​

  1. Log in to Cequence AI Gateway.
  2. Choose your tenant.
  3. Go to App catalogue.
  4. Filter by Remote MCP server.
  5. Search for Sentry MCP and then select it.
  6. Click Create MCP server.
  7. Choose auth method. If OAuth, you need an auth app with client credentials in your vendor account (see Prerequisites). Sentry supports DCR, so OAuth setup is usually straightforward.
  8. Complete the setup as prompted, select tools, and deploy.

Use the generated MCP server URL in your client as described in the Client Configuration docs. For detailed UI steps and screenshots, see Create a third-party MCP Server.

7. Additional information​

  • Seer integration: MCP and Seer are complementary—MCP brings Sentry context into your LLM; Seer is used for deep issue analysis and automated debugging. You can use MCP to invoke Seer for complex workflows.
  • Verified clients: Sentry MCP has been verified with Claude (Desktop and Claude.ai), Claude Code, Codex, Cursor, Factory Droid, Amp, VS Code/GitHub Copilot, Vercel v0, Warp, Windsurf, and other clients that support OAuth and remote MCP.
  • Troubleshooting: If OAuth fails, try the legacy Remote MCP configuration; ensure the URL is https://mcp.sentry.dev/mcp (or https://mcp.sentry.dev/sse for SSE). Clear cached auth if needed (e.g. ~/.mcp-auth).
  • Official documentation: Sentry MCP Server, Live demo, GitHub.