Sentry MCP Server

Sentry

Sentry is an application monitoring platform for error tracking, performance monitoring, and release management. This bundle provides access to your Sentry organization's projects, issues, alerts, and release data.

Entity Hierarchy

• Organization is the top-level boundary. All resources are scoped to an organization slug.
• Teams group members and own projects. A team belongs to one organization.
• Projects represent individual applications or services. Each project has its own issue stream, releases, and alert rules.
• Issues are grouped error events. Each issue has events, tags, and assignment state.
• Events are individual error occurrences within an issue, containing stack traces, breadcrumbs, and context.
• Releases track deployed versions with commit data, deploys, and associated issues.

Capabilities

Read operations:

• List and search issues with filters (project, query, sort, date range)
• Get issue details, events, and tags
• List projects and their settings
• List organizations and teams
• Get release details and deploy history
• View alert rules and their status
• Query performance data (transactions, spans)

Write operations:

• Resolve, ignore, or assign issues
• Create and update alert rules
• Manage project settings
• Create releases and associate commits

Workflows

Triage errors: List recent unresolved issues sorted by frequency or last seen. Drill into an issue to see the latest event's stack trace and breadcrumbs. Assign to a team member or resolve.

Monitor a release: Look up a release by version to see new issues introduced, crash-free session rate, and deploy history. Compare against the previous release.

Set up alerting: Create alert rules on a project scoped to specific error types, frequency thresholds, or performance metrics. Route notifications to the appropriate team.

Investigate performance: Query transaction data to identify slow endpoints. Drill into spans to find database queries or external calls causing latency.

Gotchas

• Organization slug required: Most operations need the organization slug. List organizations first if you don't have it.
• Issue vs event: An issue is a group of similar events. Searching issues returns the group; you need a separate call to get individual event stack traces.
• Pagination: Issue and event lists are paginated. Use cursor-based pagination to iterate through large result sets.
• Rate limits: Sentry enforces per-organization API rate limits. Batch operations may need throttling.