The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes a 'Companion check' block that instructs the agent to run the ls command on local skill directories (e.g., ~/.claude/skills/) to detect which related tools are already installed.- [EXTERNAL_DOWNLOADS]: The skill facilitates the installation of complementary SEO tools from the AbsolutelySkilled/AbsolutelySkilled collection using the npx skills add command.- [INDIRECT_PROMPT_INJECTION]: The skill acts on data from external websites (titles, meta descriptions, and content body) which constitutes untrusted input.
Ingestion points: Website metadata (SKILL.md) and content analyzed during checks O1-O8 and C1-C7.
Boundary markers: Not present; the skill instructions do not specify using delimiters to separate website content from agent instructions.
Capability inventory: The skill's functions are limited to auditing and report generation; it does not possess capabilities for dangerous file system writes or unauthorized network data transmission.
Sanitization: Not present; there are no instructions for sanitizing or validating external content before it is processed by the model.

seo-audit