Skip to content

Security: TylerSteeves/aibutler

Security

SECURITY.md

Security Policy

aiButler can hold secrets and perform supervised actions on a user's machine. Security is part of the product, not a cleanup task.

Reporting

Until a dedicated disclosure inbox exists, report security issues privately to tyler@aibutler.me. Do not open public GitHub issues for vulnerabilities involving auth bypass, bridge exposure, secret leakage, or silent privilege escalation.

Core rules

  • remote surfaces must never enable full-access
  • bridge access must be authenticated when exposed beyond loopback
  • secrets should live in OS credential stores where possible
  • risky actions must respect approvals and leave receipts

Current boundaries

  • the bridge is intended for trusted local-network pairing, not internet exposure
  • desktop onboarding is ahead of the mobile UX
  • this local workspace contains experimental material that should not be included in the public repo export

There aren't any published security advisories