brand-visual

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

• Third-party content exposure detected (high risk: 0.90). The skill's web-design-codex explicitly instructs the agent to fetch and "Apply all rules" from an external public URL (https://raw.githubusercontent.com/vercel-labs/web-interface-guidelines/main/command.md) as part of its mandatory workflow (see examples/skills/web-design-codex.md and references/skills/web-design-codex.md), which causes the agent to ingest untrusted third-party content that can materially determine subsequent checks and actions.

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

• Potentially malicious external URL detected (high risk: 1.00). The web-design-codex mini-skill explicitly requires fetching https://raw.githubusercontent.com/vercel-labs/web-interface-guidelines/main/command.md at runtime (via WebFetch) and uses the fetched document as the authoritative rules and output-format instructions that directly control the agent's review behavior and outputs.