The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill uses the npx skills CLI to discover and download packages from the open agent skills ecosystem (skills.sh). It references external repositories from established organizations as examples of high-quality sources.
[REMOTE_CODE_EXECUTION]: The skill provides instructions for installing and executing external code via the npx skills add command. To mitigate risk, the instructions explicitly mandate a verification process where the agent must check install counts, source reputation (favoring official vendors), and GitHub stars before recommending or performing an installation.
[PROMPT_INJECTION]: The skill involves processing external data (search results from the CLI) which could potentially contain malicious instructions (Indirect Prompt Injection).
Ingestion points: Data returned from the npx skills find command in SKILL.md (Step 3).
Boundary markers: None identified.
Capability inventory: Capability to install and execute code globally via npx skills add -g -y (Step 6).
Sanitization: The skill provides an internal verification framework (Step 4) requiring the agent to audit the reputation and metrics of any found skill before interacting with it.

find-skills