The Agent Skills Directory

[PROMPT_INJECTION]: The skill architecture includes a multi-step interview process to gather user data, which is then interpolated into generated legal documents and code. This creates a surface for indirect prompt injection.
Ingestion points: User responses collected during the interview defined in ./scripts/interview.md.
Boundary markers: No specific delimiters or instructions to disregard embedded commands were found in the skill's logic.
Capability inventory: The skill generates and writes various file types including .mdx (legal content) and .tsx (React components for UI).
Sanitization: The skill does not explicitly detail any filtering or validation of user-provided strings before interpolation.
[EXTERNAL_DOWNLOADS]: The skill includes an informational link to the developer's Discord community for support and feedback.
URL: https://discord.gg/wQWpEpnBfE in the completion message.

privacy-global-jp