The Agent Skills Directory

[SAFE]: The skill implements standard website auditing logic, including crawling and scraping metadata for SEO analysis. All operations are confined to the intended purpose of site auditing.
[PROMPT_INJECTION]: The skill processes untrusted content from external websites via the browser_crawl and browser_scrape tools to generate reports. While this establishes a surface for indirect prompt injection where a malicious site could attempt to influence the agent's output, the risk is inherent to the tool's purpose and is assessed as low severity.
Ingestion points: browser_crawl and browser_scrape calls in SKILL.md.
Boundary markers: None specified.
Capability inventory: Reads web content and writes markdown reports to the local filesystem.
Sanitization: Not explicitly defined in the instructions.

site-auditor