gamma-mind-meld-ops

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

• Third-party content exposure detected (high risk: 0.70). Runtime LLM context can include outsider-authored free text via --prompt "$PROMPT" in scripts/mind_meld_smoke.sh, where PROMPT is user-supplied (not authored by the operating user/organization) and is passed into tools/run_mind_meld_cli.py for the model.