The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The installation instructions in SKILL.md involve cloning a repository from the author's GitHub account (github.com/thatrebeccarae/claude-marketing.git). This is a standard and transparent distribution method for this skill and points directly to the developer's infrastructure.
[PROMPT_INJECTION]: The skill operates by analyzing user-provided marketing content, which introduces a surface for indirect prompt injection.
Ingestion points: User prompts containing landing page text, form fields, and funnel descriptions in SKILL.md and EXAMPLES.md contexts.
Boundary markers: No explicit delimiters are used to separate user data from the audit instructions.
Capability inventory: The skill is strictly instructional and text-based; it lacks the ability to execute shell commands, perform file system operations, or initiate network requests (except via listed integrations like Google Analytics, which are standard for the domain).
Sanitization: No explicit sanitization of input content is defined, but the lack of executable capabilities mitigates the risk.

cro-auditor