The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill facilitates the discovery and installation of external agent skills from the skills.sh registry and GitHub repositories.
[REMOTE_CODE_EXECUTION]: The core purpose of the skill is to install and execute third-party code. To mitigate risks, it provides the agent with verification criteria, such as checking for high install counts and prioritizing official sources from well-known organizations.
[COMMAND_EXECUTION]: The skill utilizes shell commands (npx skills find, npx skills add) to manage the agent's extended capabilities.
[INDIRECT_PROMPT_INJECTION]: The skill ingests data from external search results.
Ingestion points: Search results from the npx skills find command.
Boundary markers: None explicitly defined for the search output.
Capability inventory: Shell command execution via the npx skills add command.
Sanitization: The instructions mandate a manual verification step (Step 4) where the agent must evaluate the source reputation and install metrics of a skill before presenting it to the user.

find-skills