The Agent Skills Directory

[COMMAND_EXECUTION]: The skill framework references the use of curl and web_fetch to retrieve the source code of websites for technical SEO analysis. This functionality is a standard requirement for the skill's stated purpose and does not involve executing untrusted scripts or commands.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes content from external websites which could contain malicious instructions.
Ingestion points: External website source code and metadata fetched via web_fetch or curl, as well as local context from .claude/product-marketing-context.md.
Boundary markers: Absent; the instructions do not specify delimiters or guidelines to distinguish untrusted web content from the agent's core instructions.
Capability inventory: The agent's capabilities in this context are limited to text analysis, SEO auditing, and generating recommendations.
Sanitization: No explicit sanitization or validation of the fetched HTML/text content is described in the audit framework.
[DATA_EXPOSURE]: The skill instructs the agent to check for the existence of and read .claude/product-marketing-context.md. This is a localized context-gathering operation intended to improve the relevance of the SEO audit and does not target sensitive system configurations, SSH keys, or environment variables.

seo-audit