The Agent Skills Directory

[SAFE]: The skill acts as a navigation layer for other tools and provides guidance on session management. It does not contain executable code, dangerous commands, or prompt injection patterns.
[EXTERNAL_DOWNLOADS]: The skill references a documentation link at aihero.dev to explain technical concepts like the 'smart zone'. This is a standard informational reference.
[CREDENTIALS_UNSAFE]: The skill description mentions the management of .env files and GitHub secrets in the context of a setup workflow (/wizard). These are described as standard administrative tasks for a human-in-the-loop and do not involve hardcoded secrets or unauthorized access attempts.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to read other skill definitions (SKILL.md files) to provide routing advice. While this involves processing external files, it is the primary intended function of the router and does not present an immediate security risk in this context.

ask-matt