The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to run shell commands using npx, specifically npx skills find, npx skills add, and npx skills init to manage external capabilities.
[EXTERNAL_DOWNLOADS]: The skill facilitates the download of external code from GitHub and other repositories via the npx skills add command.
[REMOTE_CODE_EXECUTION]: The use of npx skills add <owner/repo@skill> -g -y allows for the automated installation and potential execution of remote code. The -y flag is used to skip user confirmation prompts, which increases the risk of executing malicious content if the agent selects an untrusted or typosquatted repository during the discovery process.
[PROMPT_INJECTION]: The skill exposes an indirect injection surface by processing and acting upon untrusted data from an external registry.
Ingestion points: Results from the npx skills find [query] command, which includes package names and descriptions from a public registry, are ingested into the agent context.
Boundary markers: There are no boundary markers or instructions to ignore embedded commands within the search results before they are displayed or used for further actions.
Capability inventory: The agent has the capability to execute shell commands (npx) and install remote code based on these search results.
Sanitization: There is no evidence of sanitization or validation of the external content retrieved from the skills.sh registry or GitHub descriptions.

find-skills