wacli

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

• Third-party content exposure detected (high risk: 0.80). The SKILL.md explicitly states the skill can "send/receive messages, media, groups, status, contacts via WhatsApp Web," which means the agent would ingest user-generated, untrusted third-party messages from WhatsApp as part of its runtime behavior (receiving messages) and those messages could influence subsequent actions.