The Agent Skills Directory

[PROMPT_INJECTION]: The skill instructions define an audit process that involves reading untrusted external data, creating a surface for indirect prompt injection.
Ingestion points: The skill reads local context from files such as .agents/product-marketing.md and instructs the agent to analyze the content of external URLs provided by the user using web retrieval tools.
Boundary markers: There are no explicit instructions or delimiters defined to prevent the agent from following instructions potentially embedded within the marketing context files or the website content being audited.
Capability inventory: The skill is intended to use network tools (mentioned as web_fetch and curl) to retrieve data. While its primary purpose is reporting, the ingestion of external content into the prompt context is an inherent risk factor for prompt redirection.
Sanitization: No sanitization or filtering logic is provided for the data retrieved from external sources.

seo-audit