The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill facilitates the installation of remote code via the npx skills add command. This process downloads and executes scripts from external repositories. The use of the -y flag is notable as it allows for automated code execution by bypassing confirmation prompts.
[COMMAND_EXECUTION]: The skill executes shell commands using npx to manage, search, and update extensions within the agent's environment.
[EXTERNAL_DOWNLOADS]: The skill connects to the https://skills.sh/ registry and identifies external repositories on GitHub to retrieve new tools and configurations. It neutrally references resources from well-known sources like Vercel Labs.
[PROMPT_INJECTION]: The skill has an indirect prompt injection surface when handling search results from the remote registry.
Ingestion points: Dynamic output from the npx skills find command integrated into the agent's workflow.
Boundary markers: None are used to delimit untrusted registry content from instructions.
Capability inventory: Execution of arbitrary shell commands via the package manager and global skill installation.
Sanitization: Registry metadata is not sanitized or validated before being presented to the agent.

find-skills