axiom-shipping

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

• Third-party content exposure detected (high risk: 0.80). The skill's asc-mcp workflow and example invocations explicitly state the agent will read and "Respond to reviews via API" and "Respond to negative App Store reviews from Claude"—i.e., ingesting user-generated App Store reviews (public third-party content) from App Store Connect and acting on them—so untrusted external content can influence the agent's actions.