The Agent Skills Directory

[PROMPT_INJECTION]: The skill contains a surface for indirect prompt injection as it processes user-supplied feature names and assessment answers. However, there are no capabilities provided to the agent that would allow for exploitation of this surface. * Ingestion points: The feature argument and user answers to the six assessment questions in Phase 1. * Boundary markers: Absent; user responses are directly included in the final markdown output. * Capability inventory: The frontmatter and Rule R010 explicitly prohibit file writes and tool usage, and there are no network or execution patterns found in the script. * Sanitization: Absent.
[EXTERNAL_DOWNLOADS]: The skill references an official GitHub repository (garrytan/gstack) for source attribution. This is a neutral documentation reference to a well-known service and does not involve any runtime downloads, package installations, or remote code execution.

product-strategy