unocss

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

• Third-party content exposure detected (high risk: 0.70). The skill's extracting docs explicitly show an inline extractor example that performs async fetch('https://example.com').text() (references/core-extracting.md), which means the agent can be configured to fetch and ingest arbitrary public web content for extraction and therefore could read untrusted third-party content.