design-md

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

• Third-party content exposure detected (high risk: 1.00). This skill's URL Mode and Brand Mode explicitly instruct the agent to fetch and ingest public website content (see "Fetch the primary URL — Use web_fetch..." in SKILL.md and "Research the brand's public identity via web_fetch" in references/brand-research-protocol.md) and to extract CSS, fonts, and tokens that directly drive DESIGN.md generation and downstream enforcement rules, so untrusted third‑party page content can materially influence agent behavior.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill's URL/Brand modes explicitly call runtime fetches (e.g., "web_fetch https://brand.com" and fetching Google Fonts like "https://fonts.googleapis.com/css2?family=Inter...") and then parse that remote HTML/CSS to extract tokens which are injected into the DESIGN.md generation, so external content fetched at runtime directly controls the agent's prompts/output and is required for those modes.