The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill provides installation commands for 92 different UX pattern sub-skills hosted on an external GitHub repository (github.com/thedaviddias/ux-patterns-for-developers). These resources originate from an unverified third-party account.
[REMOTE_CODE_EXECUTION]: By instructing the agent to install external skills from an unverified source using the npx skills add command, the skill enables the potential execution of remote instructions or scripts contained within those sub-skills.
[COMMAND_EXECUTION]: The file references/categories.md contains numerous shell commands (npx skills add https://github.com/thedaviddias/ux-patterns-for-developers --skill [name]) intended for the agent to execute. This pattern is used to dynamically extend the agent's capabilities with external, unvetted components.
[INDIRECT_PROMPT_INJECTION]: The skill references an external site index (https://uxpatterns.dev/llms-full.txt) and numerous external pattern descriptions. These external sources could be modified by an attacker to inject malicious instructions into the agent's context.
Ingestion points: External text file llms-full.txt and the remote sub-skills hosted on GitHub.
Boundary markers: No delimiters or safety instructions are provided to the agent to treat this external data as untrusted.
Capability inventory: The skill facilitates the installation of additional tools and instructions via the npx skills add mechanism.
Sanitization: There is no evidence of sanitization, validation, or filtering of the external content before it is processed or installed.

ux-patterns