The Agent Skills Directory

[COMMAND_EXECUTION]: The skill mentions using external CLI tools like lighthouse and curl for performance testing in its reference documentation. These are standard developer tools used for their intended purpose of website analysis and do not involve suspicious parameters or hidden execution.
[REMOTE_CODE_EXECUTION]: No remote code execution was detected. The skill references well-known external APIs (Google PageSpeed Insights) and extensions (DataForSEO, Gemini), which are handled as legitimate service integrations.
[PROMPT_INJECTION]: The instructions do not contain any attempt to bypass safety filters or override agent behavior. The 'Quality Gates' section uses restrictive logic to ensure output quality and compliance with SEO guidelines rather than manipulating the LLM's core instructions.
[DATA_EXFILTRATION]: There is no evidence of sensitive data access or unauthorized exfiltration. Network operations are limited to standard SEO audit tasks (crawling public URLs) and accessing well-known developer APIs.

seo