Skip to main content

​
Overview

Sentry is a one-click connector that gives Rootly AI read access to your Sentry organization during investigations. When an incident is tied to a spike in a specific exception, Rootly AI can pull the error signature, the stack trace at the point of failure, and how many users are affected — automatically, without a responder digging through Sentry. Every query is read-only. Rootly AI can inspect issues, events, and user impact but cannot resolve, assign, or modify anything on the Sentry side.

​
Before You Start

  • A Sentry organization with permission to authorize third-party OAuth apps.
  • Permission to install the Rootly AI Sentry integration for that organization.

​
Setup

Open Configuration → AI → Connectors and click Connect on the Sentry card. Sign in to Sentry and authorize Rootly AI for the organization you want to connect. When the flow completes, the card flips to Connected.

​
What Rootly AI Can Read

  • Errors — active issues in the connected organization, grouped by exception signature.
  • Stack traces — the frames captured at the point of failure, with file, line, and function context.
  • User impact — how many users are affected and how frequently.
  • Recent events — events fired against a given issue in a recent time window.
Rootly AI leans on Sentry heavily during “what’s actually broken?” investigations. Pairing it with GitHub (for the recent code change) and an infrastructure connector like AWS or Cloudflare (for the underlying state) covers most product-tier incidents.

​
During an Incident

“This is the exact stack trace, and it’s hit 1,400 users so far.” Rootly AI pulls the current top exception on the affected service, along with the frame-by-frame stack trace at the point of failure and the user-impact count. Responders don’t have to open Sentry — the stack lands in the investigation output alongside the incident summary.

​
Best Practices

  • Authorize at the organization level, not per-project. Restricting the connector to one Sentry project starves investigations that span multiple services.
  • Keep Sentry project ownership assignments up to date. Rootly AI can surface who owns an affected project — a stale ownership map leads to stale owner suggestions.
  • Pair with a code connector. Sentry tells you what is failing; a code connector (GitHub) tells you what changed just before. Together they answer most “what caused this?” questions in one investigation.

​
Troubleshooting

The authorizing Sentry user or installation doesn’t have access to those projects. Confirm the installation covers every project Rootly AI should read, then disconnect and reconnect to refresh the OAuth grant.
Sentry sometimes surfaces recurring issues that share a stack signature. If the wrong issue is being cited, look at the event timestamps in Sentry directly — the connector reads what Sentry serves, so a stale grouping on the Sentry side propagates.
If a Sentry admin revokes the Rootly AI installation from Sentry’s side, Rootly’s card still shows Connected but every query fails. Reconnect from Rootly to reinstall.

​
Frequently Asked Questions

No. Every connector is read-only. Rootly AI can read Sentry data but never invoke write actions.
Coverage depends on how your Sentry Self-Hosted instance is reachable. If it’s public HTTPS with OAuth exposed, contact support to confirm compatibility.
Rootly does not maintain a persistent copy of Sentry data — queries run at investigation time and the response is used in reasoning only. Query results do appear in the LLM traces Rootly logs for quality monitoring, as with every tool call Rootly AI makes. See Data Privacy for Rootly AI for the retention boundary.

Connectors Overview

All connectors and how they fit together.

GitHub

Pair Sentry with GitHub for the strongest “what caused this?” signal.

Data Privacy for Rootly AI

What Rootly AI sees, retention, and model training controls.