ai-incident-postmortem-template

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

• Third-party content exposure detected (medium risk: 0.65). The required workflow ingests multiple evidence-pack entries and incident timeline “reconstructed from the evidence pack” (Step 4/3), and evidence packs commonly include outsider-authored free text such as customer tickets, scribe notes, or other third-party communications; if those are included as readable prose, they can be fed into the LLM context via the evidence pack reading step.