The Agent Skills Directory

[COMMAND_EXECUTION]: Utilizes the npx skills command line tool to search, add, check, and update agent capabilities.
[COMMAND_EXECUTION]: Recommends global installation (-g) of packages, which typically affects the user's system environment and may require elevated permissions.
[EXTERNAL_DOWNLOADS]: Fetches skill data and package information from the skills.sh registry and external GitHub repositories.
[EXTERNAL_DOWNLOADS]: Directs the agent to utilize trusted resources such as the Vercel Labs official agent skills repository.
[REMOTE_CODE_EXECUTION]: Provides instructions to download and execute remote code via npx skills add with the -y flag, which bypasses confirmation prompts during installation.
[PROMPT_INJECTION]: Vulnerable to indirect prompt injection as it ingests and processes search results from the third-party skills.sh registry.
Ingestion points: Output from the npx skills find command as described in SKILL.md.
Boundary markers: No specific delimiters or safety warnings are provided for handling results from the external registry.
Capability inventory: Ability to execute shell commands, install software, and modify the agent's environment.
Sanitization: The skill does not define methods for sanitizing or validating descriptions and instructions retrieved from external package metadata.

find-skills