The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill instructions direct users to download and install components from a remote GitHub repository (SeanJ1ang/design-judge-skills) using the npx skills command.
[COMMAND_EXECUTION]: The documentation includes several shell commands that the user or agent is expected to execute, including npx skills add for installation and python for running test scripts.
[REMOTE_CODE_EXECUTION]: The npx skills add command fetches and executes code from a remote third-party source to install the skill components.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process user-provided design materials such as briefs, research, and technical specifications. This ingestion of untrusted content represents a vulnerability surface where malicious instructions could be embedded.
Ingestion points: User-provided materials (briefs, research, specs, images) processed by design-evaluation and design-information-prep.
Boundary markers: None explicitly mentioned in the documentation.
Capability inventory: Shell command execution (npx, python) and network access (web scraping).
Sanitization: No explicit mention of input validation or sanitization for external materials.

design-judge-skills