Skip to main content
Connect Sentry when customer complaints need to be matched to what broke: the error behind a wave of tickets, when it started, which release introduced it, and whether it is still happening. The Rippit agent searches Sentry issues and events, reads stack traces, and cites them in its answer.

What the agent gets

  • Issue and event search across the projects you can access, filtered by project, environment, release, and time.
  • One issue’s details, including the stack trace, tags, and the events behind it.
  • Traces, so the agent can follow an error across services.
  • Organizations and projects, so the agent can find where an error lives.
Sentry adds tools over time. A connection picks up new tools when it reconnects, so Tool permissions on the server’s page is the authoritative list for your connection.

Setup

In Rippit, open Connections → MCP Servers, choose Sentry, and sign in to Sentry to approve Rippit. Rippit registers itself automatically, so there is no client ID, client secret, or API token to create.

Identity and scope

A Sentry connection acts as the Sentry user who signed in and can reach the organizations and projects that user can access. Sentry asks for write permissions on its consent screen because its server offers write tools; Rippit only enables the tools you allow. Sentry SaaS (sentry.io) is supported. Self-hosted Sentry is not supported.

Tools

You choose which tools the agent can use under Tool permissions on the server’s page, one at a time or a whole group at once. See Setting one up. The connection starts read-only: only tools Sentry declares read-only are enabled. Sentry also offers tools that change your account, such as resolving or assigning issues and creating projects, teams, and DSNs. Turn those on in Tool permissions only if you want the agent to take those actions.