The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill facilitates the installation and execution of external code using the npx skills package manager. It specifically instructs the agent to use the -y flag, which bypasses confirmation prompts during the installation of third-party packages from the ecosystem.
[COMMAND_EXECUTION]: The skill relies on executing shell commands, including npx skills find and npx skills add, to search for and modify the local agent environment.
[EXTERNAL_DOWNLOADS]: Fetches skill configurations, metadata, and code from the skills.sh registry and GitHub repositories. It explicitly recommends sources such as vercel-labs/agent-skills, which is a trusted organization.
[PROMPT_INJECTION]: The skill contains a vulnerability surface for indirect prompt injection.
Ingestion points: The skill processes and displays output from the npx skills find command, which retrieves unvalidated descriptions and metadata from a public registry.
Boundary markers: There are no explicit delimiters or instructions provided to the agent to disregard potentially malicious instructions embedded within the search results or skill descriptions.
Capability inventory: The skill possesses significant capabilities, including the ability to install and execute arbitrary remote packages via npx.
Sanitization: There is no evidence of sanitization or verification of the content returned from the external registry before it is presented to the agent or used to trigger installation commands.

find-skills