OpenAI Codex security directory

PostHog

Published by PostHog Inc.

High risk

Connects your assistant to PostHog product analytics and management tools. Query behavior, create insights, investigate errors and inspect experiments or surveys. Manage feature flags and explore dashboards, session replay, logs and LLM analytics.

Security assessment

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

  • Feature-flag actions can change production rollout behavior: Can deploy code, applications, containers, workflows, or infrastructure to a persistent environment.
  • Hosted tools include destructive cleanup actions: Can delete records, files, projects, messages, or resources that may be difficult to recover.
  • Prompts and queries are sent to PostHog's hosted service: Can send prompts, files, queries, or connected-service data to an external service.
  • The app creates and updates PostHog product objects: Can create or change persistent records, files, projects, permissions, settings, or credentials.
  • The app reads product analytics, people, errors, logs and replays: Can read private account, workspace, repository, message, analytics, or file data.
  • PostHog MCP executes arbitrary HogQL and SQL queries: Can execute code, commands, scripts, unrestricted queries, or dynamic gateway actions.
  • Some hosted MCP tools can consume billed PostHog AI spend: Can initiate purchases, paid jobs, payments, refunds, infrastructure spend, or other financial activity.

Available capabilities

This add-on exposes 1 tools or capabilities.

  • posthog

The interactive security report will load automatically.