design-web-guidelines

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

• Third-party content exposure detected (high risk: 1.00). The skill instructs the agent to use WebFetch to retrieve and apply the latest guidelines from a public raw GitHub URL (https://raw.githubusercontent.com/vercel-labs/web-interface-guidelines/main/command.md), meaning it will fetch and interpret untrusted, user-editable third-party content at runtime which could carry indirect prompt injection.

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

• Potentially malicious external URL detected (high risk: 1.00). The skill explicitly requires fetching runtime guidelines from https://raw.githubusercontent.com/vercel-labs/web-interface-guidelines/main/command.md, and the fetched document "contains all the rules and output format instructions" that directly control the agent's prompts/behavior and are required for the review, so this is a remote content dependency that controls execution.