The Agent Skills Directory

[SAFE]: The skill is entirely documentation-based, consisting of markdown guides, frameworks, and templates. There are no executable scripts or binary files included.
[SAFE]: Security best practices are followed in the instructions, which mandate that the agent must preserve parent safety and approval rules for any destructive or security-sensitive tasks.
[SAFE]: The metadata configuration (.skillmeta.yml) implements path restrictions for upstream contributions, preventing the accidental leakage of local configuration files, user data, or company secrets.
[SAFE]: While the skill involves the analysis of external content (e.g., SEO data and competitor pages), it does not provide any autonomous tools for network access or file modification. It relies on the host environment's existing safety guardrails to process this information securely.

content-audit