The Agent Skills Directory

[PROMPT_INJECTION] (LOW): Indirect Prompt Injection surface detected. The skill processes untrusted user input and existing project files into documentation.
Ingestion points: User responses from AskUserQuestion and project data from docs/PRD.md.
Boundary markers: Absent; content is interpolated directly into templates.
Capability inventory: Read, Write, Edit, Glob, Grep, Bash, and AskUserQuestion.
Sanitization: None detected.
[COMMAND_EXECUTION] (SAFE): The Bash tool is used only for git ls-files to inspect the project structure, which is a benign and necessary operation for the skill's workflow.
[DATA_EXFILTRATION] (SAFE): No network tools are enabled, and no patterns for exfiltrating sensitive data to external domains were found.

requirements