The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection as it processes untrusted image URLs provided by the user.
Ingestion points: The person_image and clothing_image inputs in SKILL.md are the entry points for untrusted data.
Boundary markers: There are no delimiters or instructions to ignore embedded content when these inputs are interpolated into tool prompts.
Capability inventory: The skill uses the muapi CLI and curl for network-based image and video generation, representing the available capabilities.
Sanitization: The skill does not mention any validation, filtering, or sanitization of the external content or URLs.
[COMMAND_EXECUTION]: The skill utilizes the muapi CLI and curl commands to perform its primary functions. It correctly manages secrets by instructing the agent to use the MUAPI_API_KEY environment variable rather than hardcoding sensitive credentials.

muapi-fashion-try-on