find-skills

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 1.00). The skill explicitly instructs using npx skills find/add and links to public sources (e.g., "npx skills find", "npx skills add" and https://skills.sh/ and GitHub) so the agent will fetch and read community-published Skills from open public sites and use those results to decide which Skills to install, which could allow untrusted third-party content to influence tool use.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill instructs at runtime to install external Skills from GitHub (e.g., "npx skills add vercel-labs/agent-skills@vercel-react-best-practices" -> https://github.com/vercel-labs/agent-skills), which fetches and installs remote skill code/definitions that can directly control agent prompts or execute code.