agent-patterns

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

• Third-party content exposure detected (high risk: 0.80). The skill explicitly grants and recommends subagents use WebFetch/WebSearch and instructs workflows to "Verify with [external check]" and "Check [authoritative source]" (see "Tool Access Patterns" and delegation/workflow templates in SKILL.md and references/*.md), which means the agent will fetch and interpret arbitrary public web content that can materially influence actions.