wireframe-skill

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

• Third-party content exposure detected (medium risk: 0.65). The required workflow is installing and invoking the community skill yhassy/wireframe-skill from a public GitHub repo; at runtime the agent will execute that outsider-authored code, which can ingest outsider-authored free text (e.g., user-provided natural-language UI descriptions and any fetched/embedded content) into the LLM context via the skill’s implementation.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill explicitly instructs executing an npx install (e.g., "npx skills add yhassy/wireframe-skill") which at runtime will fetch and run remote package code from the upstream repository (https://github.com/yhassy/wireframe-skill), so external content is fetched and executed during installation.