production-agent-public

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

• Third-party content exposure detected (high risk: 1.00). SKILL.md explicitly requires using web tools like "tool request browse_page with url" and the examples (references/示例输出.md showing fetcher.py with TARGET_URL "https://example.com/data") demonstrate fetching and scraping external web pages, meaning the agent will ingest untrusted public web content that can influence its subsequent actions.