The Agent Skills Directory

[SAFE]: The skill is entirely instructional, focused on providing an architectural review framework for implementation plans. It does not perform network requests, file writes, or command execution.
[SAFE]: It promotes security best practices by requiring a specific security/privacy check for plans involving sensitive areas like sign-in, permissions, secrets, or user data.
[SAFE]: References to other skills (e.g., $golang-pro, $vercel-react-best-practices) are standard platform cross-skill consultations and do not involve untrusted remote code or third-party package installations.
[SAFE]: The skill uses structured output formatting and specific rules for analysis, which serves as a defensive measure against following instructions that might be embedded in the reviewed plans.

sensei-gameplan