The Agent Skills Directory

[COMMAND_EXECUTION]: The skill is designed to wrap and execute the npx skills command-line utility. It provides specific instructions for searching (find), checking for updates (check), and performing updates (update) on the local system.
[EXTERNAL_DOWNLOADS]: The skill facilitates the installation of remote code via the npx skills add <package> command. This process downloads modular packages from external repositories, such as GitHub. The instructions explicitly direct the agent to use the -y flag to skip user confirmation prompts and the -g flag for global installation, which grants the installed code persistent access to the user's environment.
[INDIRECT_PROMPT_INJECTION]: The skill acts as an ingestion point for untrusted data from an external registry.
- Ingestion points: The agent reads and processes output from the npx skills find command, which includes descriptions and metadata for third-party skills retrieved from skills.sh.
- Boundary markers: The instructions do not define boundary markers or safety guidelines to prevent the agent from following malicious instructions that might be embedded in the descriptions of discovered skills.
- Capability inventory: The skill possesses the capability to modify the local environment and execute new code through the add command.
- Sanitization: There is no evidence of sanitization or validation of the metadata returned by the search command before it is presented to the agent or the user.

find-skills