academic-pipeline

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

• Third-party content exposure detected (high risk: 0.85). The required runtime workflow can ingest outsider-authored free text via the Stage 4→5 claim-faithfulness audit: it retrieves full-text from literature_corpus[] (public web / external PDFs) and passes retrieved excerpts into the LLM judge context for claim_ref_alignment_audit_agent (Step 2–5), which is outsider-authored content.