claude-code-academic-workflow

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

• Third-party content exposure detected (medium risk: 0.65). The workflow’s runtime LLM context can ingest outsider-authored free text from the user-provided manuscript/slides files (e.g., /review-paper manuscript.tex, /verify-claims manuscript.tex, /audit-reproducibility manuscript.tex read and quote their contents), which may include text authored by someone other than the operating user.