web-design-guidelines

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

• Third-party content exposure detected (high risk: 0.85). The skill’s required workflow fetches public web content at runtime from the specified URL (https://raw.githubusercontent.com/.../command.md) via WebFetch, and that fetched free-text guidelines are then ingested into the agent’s LLM context to drive the audit rules.

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

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