Automation sometimes unable to find sentry tools despite sentry MCP added to automation

Where does the bug appear (feature/product)?

Background Agent (GitHub, Slack, Web, Linear)

Describe the Bug

I can’t get the Sentry MCP to work with Automations.

I have the Sentry MCP configured via https://cursor.com/dashboard/integrations. I’ve confirmed that background agents I manually create in the web UI can access it and can use it to find real sentry issues in our org. However, automation runs often (but not always??) claim they cannot access Sentry. Here is an example prompt that failed five out of seven times: “First, verify sentry MCP tool access with @MCP: Sentry. We’re trying to debug some access issues, and it’s critical that you’re able to read from sentry issues to work.”

The automation test run is unable to find any sentry tools or query sentry most of the time. Any suggestions about how to reliably expose sentry access to automations?

Steps to Reproduce

  • Add the sentry MCP
  • Create an automation
  • Add the sentry MCP to the automation
  • Add something to the prompt telling the agent to search for sentry issues
  • Save
  • Trigger a test run of the automation
  • Behold the agent claim it can’t actually access sentry

Expected Behavior

The automation run should be able to access sentry via the MCP reliably.

Operating System

MacOS

Version Information

Version: 2.6.18
VSCode Version: 1.105.1
Commit: 68fbec5aed9da587d1c6a64172792f505bafa250
Date: 2026-03-10T02:01:17.430Z
Build Type: Stable
Release Track: Default
Electron: 39.6.0
Chromium: 142.0.7444.265
Node.js: 22.22.0
V8: 14.2.231.22-electron.0
OS: Darwin arm64 25.1.0

For AI issues: which model did you use?

  • Codex GPT 5.3 High
  • GPT 5.4 High

Does this stop you from using Cursor

No - Cursor works, but with this issue

Oh yeah, here is a background agent ID for a run that failed to find the Sentry MCP: bc-951dae64-9759-4417-b3aa-f33701f99295.

Here is another run where it worked: bc-da944edf-a22c-4b0d-bb73-214bb6923f53

Hey, thanks for the detailed report and the agent IDs. Both the failing and working ones are really helpful for debugging.

This is a known issue. MCP tool discovery in Automations is intermittent right now, and we’re seeing similar reports with other MCP providers too, like Datadog, New Relic, and LogFire. The team is aware and tracking it. There’s no ETA yet, but your report, especially the side-by-side agent IDs, helps us prioritize.

As a temporary workaround, some users have had better luck by retrying the automation. You mentioned it does succeed sometimes, 2 out of 7 in your case. Not ideal, I know.

We’ll update here when there’s progress. Let me know if anything changes on your end.

Hey @jamison_droplet!

This issue has been fixed in a recent Cursor update. Updating to the latest version should resolve this. If you’re still experiencing this on the latest version of Cursor, please raise a new thread and we’ll take a look.