diagnosing-root-causes

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

• Third-party content exposure detected (medium risk: 0.30). The skill’s runtime ingests outsider-authored free text directly via its required ISSUE/RESOURCES inputs into Intake → CollectEvidence → root-cause-analyst (and later review), meaning an outsider can supply logs/issues/CI text for the LLM to read and summarize into evidence.