@349/starchild-design-pack

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 is to fetch the hosted “Starchild Brand Registry” (public web content) and ingest its brand.json/tokens.css/brand.md text into the agent’s context, which is an outsider-authored source that can contain prompt-injection strings.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill explicitly tells the agent to fetch the Brand Registry at runtime (https://community.iamstarchild.com/1892-starchild-brand/) and to web_fetch the canonical guidelines (https://raw.githubusercontent.com/vercel-labs/web-interface-guidelines/main/command.md), and those external documents are required at runtime and directly control the agent's tokens/instructions used to generate outputs.