The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes shell commands using npx skills to search for, check, and update agent extensions. It includes instructions for global installations (-g) and bypassing confirmation prompts (-y).\n- [EXTERNAL_DOWNLOADS]: The skill facilitates downloading and installing external packages from GitHub and other sources via the Skills CLI. It specifically mentions using resources from well-known repositories like those maintained by Vercel Labs.\n- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes untrusted data from search results and uses it to execute installation commands.\n
Ingestion points: Data enters the agent's context through the output of the npx skills find command, which lists available packages based on a search query.\n
Boundary markers: The instructions do not define delimiters or provide warnings to the agent to ignore potentially malicious instructions embedded in skill descriptions or metadata returned by the search command.\n
Capability inventory: The agent has the capability to run shell commands and install software globally without user intervention using the -y flag.\n
Sanitization: No mechanism is described to validate or sanitize the package names or repository paths returned by the search tool before they are passed to the installation command.

find-skills