find-skills

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

• Third-party content exposure detected (high risk: 0.75). The skill’s runtime workflow calls Alibaba Cloud’s “SearchPublicMarketSkill” API (public marketplace) and then ingests the returned JSON fields like description/source_url into the agent to present install options, which is outsider-authored free text from a public web/marketplace source.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill workflow includes runtime installation commands like "npx skills add https://github.com/owner/repo -s <skill_name>" (and other repo forms such as https://gitlab.com/org/repo or git@github.com:owner/repo.git) which fetch and install remote skill code that will execute in the agent environment and can directly control agent behavior.