The Agent Skills Directory

[SAFE]: The skill consists exclusively of instructional markdown and configuration files. No executable scripts, binaries, or command-line operations were found.
[SAFE]: No network operations (e.g., curl, wget) or external resource fetches were detected. All references are internal to the skill package (e.g., local file paths).
[SAFE]: No sensitive data exposure or exfiltration patterns were identified. The skill does not access environment variables, credential files, or user secrets.
[SAFE]: The framework uses structured templates and explicit constraints to guide AI behavior, reducing the risk of unintended output without employing obfuscation or bypass techniques.
[INDIRECT_PROMPT_INJECTION]: The skill acts as an ingestion surface for user-provided prop descriptions which are then interpolated into templates for external platforms (Midjourney, Stable Diffusion).
Ingestion points: User-provided prop dimensions and descriptions in SKILL.md (Section 1).
Boundary markers: The skill uses a structured JSON interface (Section 10) and explicit dimension tables to isolate input data.
Capability inventory: None; the skill does not possess file-writing, network, or code execution capabilities.
Sanitization: Content is constrained by strict dimension rules and negative prompt filters defined in references/NEGATIVE-CASE-BOOK.md. Given the lack of dangerous capabilities, this surface is considered safe.

prop-asset