The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands via bunx or npx to interact with the skills and skvlt (Skills Vault) command-line tools. These commands are used for discovering, installing, and managing local and global agent skills.
[EXTERNAL_DOWNLOADS]: The skill facilitates the download and installation of extension code from external repositories on GitHub and GitLab, as well as direct URLs. The instructions advise the agent to perform quality checks, such as verifying install counts and source reputation (e.g., official organizations), to mitigate the risk of installing untrusted code.
[INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface where it ingests and processes potentially untrusted data from external sources during the discovery phase.
Ingestion points: Output generated by bunx skills find, repository descriptions, and skill package listings (--list).
Boundary markers: No specific delimiters or boundary instructions are provided to isolate external data from the agent's primary instructions.
Capability inventory: The skill has access to shell execution and the ability to modify the local environment through package installation.
Sanitization: There are no requirements for sanitizing or validating search results or metadata fetched from external registries before the agent acts on the information.

skills-cli