The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to process external content including live website data and project-specific blueprint files.
Ingestion points: Untrusted data enters the agent context through the website URL and the Site Blueprint file stem specified in the Routing section.
Boundary markers: The instructions lack specific delimiters or instructions to ignore potential command-like patterns within the external files or website content.
Capability inventory: The skill has the capability to trigger further automation steps (website-audit, homepage-rewrite, page-builder) based on the status and content found in these external inputs.
Sanitization: There is no evidence of sanitization or validation logic for the content ingested from the target URLs or local blueprint files.
[EXTERNAL_DOWNLOADS]: The skill refers the user to installation commands for supplementary tools required for the workflow.
Evidence: The Routing section provides commands such as npx skills add victor-shulga/website-audit/skills/website-audit to install missing dependencies.
Context: These external resources are hosted within the skill author's (victor-shulga) repository ecosystem, representing standard modular design for this toolset.

website