web-design-guidelines

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly instructs the agent to fetch and apply rules from a public URL (https://raw.githubusercontent.com/vercel-labs/web-interface-guidelines/main/command.md) using WebFetch before each review, meaning untrusted third-party content is read and can directly change the agent's checking behavior and outputs.

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

• Potentially malicious external URL detected (high risk: 1.00). The skill explicitly fetches runtime guidelines from https://raw.githubusercontent.com/vercel-labs/web-interface-guidelines/main/command.md and instructs the agent to read and apply those fetched rules and output-format instructions, so the external content is required and directly controls the agent's prompts/behavior.