agent-skills-discovery

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

• Third-party content exposure detected (low risk: 0.10). The workflow searches the web via Exa search and previews third-party GitHub repositories/skills via gh skill preview, which ingests outsider-authored free text from the web and external repositories.

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

• Potentially malicious external URL detected (high risk: 0.70). The skill contains instructions to execute curl -sSL https://cli.runpod.net | bash at runtime to install an external binary dependency from an unauthenticated, third-party hosting script.