ai-agent-safety-and-red-team

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

• Third-party content exposure detected (high risk: 0.85). The required workflow ingests outsider-authored free text via the red-team CI suite’s runtime corpus fixtures (e.g., tests/ai/red-team/agent/indirect_injection/corpus/kb_chunks, web_pages, customer_emails, and tool_responses), which are loaded into the agent as KB chunks / scraped HTML / email bodies / mocked tool outputs and thus become LLM-readable context.