The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external content from target URLs and search engine results, creating a surface for indirect prompt injection attacks where a malicious website could attempt to influence the agent's behavior. \n * Ingestion points: The skill fetches external data via render_page.py, parse_html.py, and WebSearch as part of its execution pipeline (SKILL.md).\n * Boundary markers: The instructions lack explicit delimiters or instructions to treat the fetched content as untrusted data or to ignore any commands contained within it.\n * Capability inventory: The skill utilizes local script execution via claude-seo run, web searches, and DataForSEO API integrations.\n * Sanitization: The instructions do not define any sanitization or validation steps for the content retrieved from external sources before it is analyzed by the agent.\n- [COMMAND_EXECUTION]: The skill's workflow includes executing local Python scripts (render_page.py, parse_html.py) via a custom command-line interface (claude-seo run). These scripts take user-supplied input such as URLs and keywords, representing a command execution surface.

seo-sxo