Sentry Alternative MCP. Audit errors and releases from your chat client.

Listen up. This server connects your AI client straight into Sentry’s entire error database. You don't gotta jump back to a web dashboard every time something breaks; your agent handles all the deep dives right here in the chat window. It keeps you on the fix, period.

Handling Issues and Bugs:
You can find, search, and update specific bugs across any project. You use list_issues or search_issues to pull a list of issues either organization-wide or scoped down to one spot. You can filter that dump by status—like only showing open or high-priority stuff. Once you've found the bug, you can even change its state using update_issue, assigning it to another user or tagging it as resolved.

Want a deeper dig? Running search_issues lets you use advanced Sentry query syntax (think things like finding all issues where 'is:open' and 'priority:10') across the whole organization, which is way faster than clicking through menus.

Inspecting Raw Failures:
If you need to know exactly what broke, you pull up the raw event data. You call get_event with a specific ID to grab all the details on a single error event. That includes full stacktraces—the whole chain of calls—plus platform context and breadcrumbs. For a list of recent failures, running list_events gives you quick access to basic stacktrace snippets and timestamps across a project.

If you need the entire profile history for an issue, use get_issue with its numeric Sentry Issue ID.

Mapping Errors to Code Versions:
You gotta know which version of the code caused the mess. You can track deployments using list_releases, which pulls back all past application versions (the releases) for both an entire organization or a specific project. This lets you pinpoint exactly what code change introduced the error or regression.

Auditing and Structure:
You gotta see who owns what, and how alerts are set up. You use list_organizations to get a list of all connected Sentry organizations and their unique slugs. To narrow your search, you can run list_projects to fetch every project within an organization. Need to know the teams? list_teams shows those too—all the members and what projects they're associated with.

You can also run list_tags to see all the tags used across multiple projects, which helps immensely when you need to group errors for analysis.

And don't forget about the alerts. If something goes wrong, who gets notified? Use list_alert_rules. This function lists every single configured alert rule, showing exactly what conditions trigger those notifications—whether it's a Slack message, an email blast, or some other service call. It lets you audit the entire notification system.

Getting Started:
Before anything else, your AI client verifies its credentials. You run get_auth_info to confirm that your token is active and correctly set up for the server. Just tell your agent what you need—like 'show me all open issues in the payment service' or 'list every project slug.' Your agent executes the necessary tool calls, pulling back the formatted answer with all the data you need.