The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill provides instructions for the agent to install and execute external code packages using the npx skills add command.
[EXTERNAL_DOWNLOADS]: The skill facilitates the download of content from external, unverified sources, including arbitrary GitHub repositories and the skills.sh registry.
[COMMAND_EXECUTION]: The skill executes shell commands for searching and installing packages. The recommended installation command npx skills add <package> -g -y is particularly concerning because the -y flag bypasses user confirmation prompts during the installation of remote code.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection due to its reliance on untrusted external data.
Ingestion points: Skill descriptions and metadata are fetched from the external skills.sh service via the npx skills find command.
Boundary markers: There are no boundary markers or instructions to the agent to ignore embedded commands within the search results.
Capability inventory: The agent possesses the capability to execute shell commands and perform global package installations.
Sanitization: No sanitization or verification of the content returned from the remote registry is performed before the agent processes it or offers to execute the installation.

find-skills