The Agent Skills Directory

[SAFE]: No security findings were identified. The skill adheres to its stated purpose of generating and refining marketing copy. It provides structured references for various platforms (Google Ads, Facebook, LinkedIn, etc.) and enforces character limits and tone constraints.
[INDIRECT_PROMPT_INJECTION]: The skill processes user-provided inputs such as product briefs and existing copy to generate marketing materials. While it has access to tools like Bash and Write, it lacks automated execution chains that would allow untrusted data to trigger high-risk commands. Ingestion points include the landing-copy --improve command. Capability inventory includes file system access and shell execution, but these are gated by instructions to use them only for copy manipulation and linting. Sanitization is partially addressed by a mandatory pass through a separate writer tool for typography and style cleanup.

landing-copy