auth0-branding

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly fetches and ingests content from arbitrary public websites via Brandfetch when a user supplies a URL (see references/capability-brand.md and SKILL.md) and also “samples the copy” from user-provided URLs to classify and generate voice-matched rewrites (references/capability-voice.md), meaning the agent reads untrusted third‑party content that can materially influence subsequent actions.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill explicitly calls the Brandfetch API at runtime (GET https://api.brandfetch.io/v2/brands/) to extract colors, logos, and fonts that are directly injected into the agent's proposal and subsequent tenant writes for the "brand from URL" flow, so this external content can directly control agent prompts/results for that path.