How to troubleshoot and block email leaks in Claude Code request headers

Because Claude Code places user login emails in the model context, the model can inadvertently write them into User-Agent request headers when autonomously generating network request commands, causing privacy leaks. Anthropic introduced a temporary built-in cc-plugin-privacy plugin in version 2.1.317 and above to filter this data; however, the check only covers certain shell invocations and cannot intercept MCP tools, WebFetch, or script-based writes.