The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data by fetching and extracting information from external websites using the webfetch tool to perform SEO audits.
Ingestion points: External HTML content, hreflang tags, and on-page metadata are fetched from user-provided site URLs in the 'Data pulls' section.
Boundary markers: The instructions lack explicit delimiters or warnings to the agent to disregard potential instructions embedded within the fetched SEO data.
Capability inventory: The agent can perform network operations via webfetch and execute local scripts (scripts/report_publish.py).
Sanitization: There is no mention of sanitizing or validating the content extracted from remote pages before it is processed by the agent.
[COMMAND_EXECUTION]: The skill instructs the agent to execute a local Python script to finalize its output.
Evidence: The 'Output' section requires running python scripts/report_publish.py <report>.md to generate branded deliverables. While this is a common workflow for this vendor, executing bundled scripts involves the execution of local code.

international-seo