The Agent Skills Directory

Indirect Prompt Injection (HIGH): The skill is designed to ingest and process untrusted data from external sources such as website architecture, content inventories, and competitor landscapes.
Ingestion points: Website content and metadata retrieved via google-search-console, screaming-frog, and web-reading tools.
Boundary markers: None. No instructions exist to delimit external content or ignore embedded commands.
Capability inventory: Possesses Bash, Write, and Read capabilities, which are highly exploitable.
Sanitization: None detected. The agent interpolates external data directly into its reasoning process.
Command Execution (HIGH): The inclusion of the Bash tool provides a mechanism for arbitrary command execution. When combined with the lack of input sanitization for external SEO data, an attacker could embed malicious shell commands in a website's metadata or hidden tags that the agent might execute during an audit.

seo-specialist