OpenAI Codex security directory

Sentry

Published by OpenAI

High risk

Helps inspect Sentry issues, events and production-error health data. The bundled helper performs read-only API queries and requires a Sentry auth token. Enabled triage or project/DSN administration uses the separate hosted Sentry MCP.

Security assessment

Plutonium assessed this plugin for permissions, capabilities, and security-relevant behavior.

  • Hosted MCP can create and modify Sentry projects and operational monitors: Can deploy code, applications, containers, workflows, or infrastructure to a persistent environment.
  • Can change persistent data or access: Can create or change persistent records, files, projects, permissions, settings, or credentials.
  • Reads production errors, events, stack traces and organization metadata: Can read private account, workspace, repository, message, analytics, or file data.
  • Runs AI-translated searches against private Sentry telemetry: Can execute code, commands, scripts, unrestricted queries, or dynamic gateway actions.

Available capabilities

This add-on exposes 1 tools or capabilities.

  • sentry

The interactive security report will load automatically.