mcp

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

• Third-party content exposure detected (high risk: 0.90). The skill's documentation and workflow (SKILL.md and references/patterns.md) explicitly instruct connecting to remote MCP HTTP endpoints (e.g., HttpClientTransport with Endpoint -> "https://example.com/mcp"), listing remote tools/prompts (ListPromptsAsync/ListToolsAsync), and passing discovered prompts/tools into an IChatClient—which means untrusted third-party server-provided prompts/resources can be ingested and directly influence agent behavior.