The Agent Skills Directory

[PROMPT_INJECTION]: The skill implements a logic for parsing user requests and project context to generate architectural questions. This pattern of ingesting untrusted data without specified boundary markers or sanitization procedures creates a surface for indirect prompt injection. However, the impact is minimized by the skill's restriction to read-only tools and text-based output.
Ingestion points: SKILL.md (Question Generation Process)
Boundary markers: Absent in the parsing logic
Capability inventory: Access to Read, Glob, and Grep tools (read-only)
Sanitization: No validation or sanitization of user-supplied data is described.
[NO_CODE]: The skill is composed entirely of Markdown documentation and instructional protocols. No executable scripts, binaries, or package dependencies were found in the provided files.

brainstorming