The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data (raw HTML and rendered DOM) fetched from websites. This creates a surface for indirect prompt injection where a malicious website could include instructions designed to influence the agent's findings or behavior during the audit process.
[COMMAND_EXECUTION]: The skill utilizes the Bash tool and instructions mention creating runnable assertions (render_diff) for verification. There is a potential risk if the agent incorporates unsanitized content from the audited web pages into shell commands or scripts, which could lead to command injection or unintended execution within the agent's environment.

seo-rendering