fact-check

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

• Third-party content exposure detected (medium risk: 0.30). The runtime workflow ingests outsider-authored free text from user-driven web search results (via the Web Forager CLI or ddgs) and then fetches selected URLs into Jina Reader markdown as part of “Search for supporting/contradicting evidence” and “Fetch key sources.”

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

• Potentially malicious external URL detected (high risk: 0.80). The skill instructs runtime fetching of arbitrary webpage content via Jina Reader (curl "https://r.jina.ai/https://example.com" and requests.get(f"https://r.jina.ai/{url}")), which will inject remote markdown into the agent's context and can directly control prompts.