Skip to content

Security: Rethunk-AI/ponytail

SECURITY.md

Security

Supported versions

The head of main. No release branches; a fix ships in the next commit.

Reporting

GitHub private vulnerability reporting (Security → Report a vulnerability). Do not open a public issue first.

In scope

  • A hook that lets prompt text, a project path or a config value reach a shell as anything but an inert argument.
  • A hook that writes outside the host's config directory or ~/.config/ponytail/.
  • Rule text injected into a session that differs from skills/ponytail/SKILL.md and AGENTS.md in this repository.

Out of scope

  • Vulnerabilities in the host agents themselves (Claude Code, Codex, Cursor and the rest). Disclose those to their vendors.
  • Issues that exist only in upstream code this fork removed; report anything still shared upstream at DietrichGebert/ponytail as well.

There aren't any published security advisories