The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill instructs the agent to install and execute code from external sources using the npx skills add <package> command. This allows for the execution of arbitrary scripts hosted on GitHub repositories.
[COMMAND_EXECUTION]: The instructions recommend using the -y flag with npx skills add, which explicitly skips confirmation prompts. This allows the agent to install third-party code without human oversight, increasing the risk of automated compromise.
[EXTERNAL_DOWNLOADS]: The skill performs network operations to search for and download content from the skills.sh domain and various GitHub repositories.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection through tool output poisoning.
Ingestion points: The agent ingests untrusted data from the output of the npx skills find command, which contains repository names and descriptions provided by third-party authors.
Boundary markers: There are no instructions for the agent to use delimiters or ignore instructions that might be embedded in the search results.
Capability inventory: The agent has the capability to execute commands and install persistent code via the npx skills utility.
Sanitization: The skill does not provide mechanisms to sanitize or validate the content of the search results before the agent processes them.

find-skills