building-pydantic-ai-agents

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly documents and instructs adding WebSearch and WebFetch capabilities (see "Add Capabilities to an Agent" in references/COMMON-TASKS.md and the Task Routing Table / Quick-Start Patterns) which enable the agent to fetch and read public web pages or search results (e.g., DuckDuckGo or native web search), thereby exposing it to untrusted third-party content that could carry indirect prompt-injection.