jina-reader

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

• Third-party content exposure detected (high risk: 0.95). The required workflow fetches arbitrary OUTSIDER-authored public web content at runtime via scripts/read_url.py calling https://r.jina.ai/ with a user-supplied URL, and then prints/returns the resulting Markdown into the agent’s LLM context.

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

• Potentially malicious external URL detected (high risk: 0.90). The skill makes runtime requests to https://r.jina.ai/ to fetch Markdown for arbitrary target pages and injects that fetched content into the agent context, which can carry prompt-injection instructions that directly influence the agent's behavior.