brainstorming

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

• Third-party content exposure detected (high risk: 0.90). This skill explicitly instructs sub-agents to perform WebSearch and ingest web content (see references/design-creation.md Sub-agent 1: "Use WebSearch for latest best practices" and Sub-agent 2: "Research best practices, security, performance patterns"), so untrusted public third-party content is read and used to influence design decisions and agent actions.