data-scraper-agent

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

• Third-party content exposure detected (medium risk: 0.30). The workflow ingests outsider-authored free text at runtime by scraping externally supplied public URLs/RSS/REST/HTML (e.g., fetch() in scraper/sources/* collects resp.text/JSON), then embedding the resulting item content into the LLM prompt in ai/pipeline.py via _build_prompt() and sending it to Gemini (ai/client.py).