The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill provides instructions for the agent to download and install external code packages using npx skills add <package>. This mechanism allows for the execution of third-party code from arbitrary GitHub repositories or the skills.sh registry within the agent's environment.
[COMMAND_EXECUTION]: The skill executes shell commands via the npx utility to perform operations such as searching (find), installing (add), and updating (update) packages. The use of the -y flag in the installation command (npx skills add <package> -g -y) is specifically designed to bypass interactive confirmation prompts.
[PRIVILEGE_ESCALATION]: The skill encourages the use of the -g (global) flag for installations. This attempts to install software at a system or user-wide level, which may involve broader permissions and affects the environment across different sessions and projects.
[INDIRECT_PROMPT_INJECTION]: The skill possesses a vulnerability surface where it ingests data from an external registry during the search process, creating an attack surface for indirect prompt injection.
Ingestion points: Data retrieved from the network via npx skills find [query] (referenced in SKILL.md).
Boundary markers: None identified; the agent is directed to process and present search results without explicit safety delimiters or instructions to ignore embedded commands.
Capability inventory: Includes shell command execution, network access for downloads, and global file system writes.
Sanitization: No evidence of sanitization or validation of the external package metadata or search results before presentation or execution.

find-skills