unocss

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

• Third-party content exposure detected (high risk: 0.80). The skill's references/core-extracting.md explicitly shows inline extraction using an async fetch (e.g., async () => (await fetch('https://example.com')).text()), which means the tool can fetch and ingest arbitrary public URLs' content and use that content as part of its extraction/workflow, allowing untrusted third-party content to influence processing.