LaunchKit AI executes generated code inside E2B sandboxes. Treat provider keys, user prompts, generated files, and sandbox URLs as sensitive application data.

Please open a private security advisory on GitHub when the repository is public, or contact the maintainer directly if advisories are not yet enabled.

• Never commit real .env.local values.
• Keep OpenAI, Clerk, E2B, Inngest, and database credentials server-side.
• Review sandbox template changes carefully.
• Avoid logging secrets, full prompts with private data, or provider responses containing sensitive content.