css-first

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

• Third-party content exposure detected (high risk: 0.70). The skill's references/live-mdn-fetch.md explicitly instructs the agent to fetch and parse live data from public third‑party URLs (e.g., https://unpkg.com/web-features/data.json, https://unpkg.com/@mdn/browser-compat-data/data.json) and to open MDN pages (developer.mozilla.org) for runtime documentation/compatibility info, so the agent will read and interpret open web content.

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

• Potentially malicious external URL detected (high risk: 0.80). The skill's "Live MDN Fetch Workflow" explicitly requires runtime fetches from https://unpkg.com/web-features/data.json and https://unpkg.com/@mdn/browser-compat-data/data.json (and the MDN pages those point to), and it uses the fetched JSON/docs to determine Baseline status and to drive the agent's responses—i.e., the external content is fetched at runtime, directly controls prompts/instructions, and is relied on as a required dependency.